Rusty Bargain Used Car Sales¶

Rusty Bargain is developing an app to attract new customers, and they need to be able to quickly find out the market value of a car. We are given access to historical data: technical specifications, trim, and prices. Rusty Bargain wants a model that will predict the value of the car, while also considering the quality and speed of the prediction, and also the time required for training.

Import Data¶

In [1]:
# !pip install --user plotly_express
In [3]:
# import libraries
import pandas as pd 
import numpy as np
import time
import plotly_express as px
import plotly.graph_objects as go
from sklearn.metrics import mean_squared_error as mse
from sklearn.preprocessing import OneHotEncoder, StandardScaler
from sklearn.linear_model import LinearRegression
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import train_test_split, cross_val_score, cross_val_predict

from sklearn.experimental import enable_iterative_imputer
from sklearn.impute import SimpleImputer, IterativeImputer, KNNImputer
from sklearn.preprocessing import OrdinalEncoder, LabelEncoder
from sklearn.pipeline import Pipeline
import lightgbm as lgb
from catboost import CatBoostRegressor, Pool 
import xgboost as xgb
In [4]:
# read dataset
df = pd.read_csv('datasets/car_data.csv', parse_dates=True)
In [5]:
# look at data
df.head()
Out[5]:
DateCrawled Price VehicleType RegistrationYear Gearbox Power Model Mileage RegistrationMonth FuelType Brand NotRepaired DateCreated NumberOfPictures PostalCode LastSeen
0 24/03/2016 11:52 480 NaN 1993 manual 0 golf 150000 0 petrol volkswagen NaN 24/03/2016 00:00 0 70435 07/04/2016 03:16
1 24/03/2016 10:58 18300 coupe 2011 manual 190 NaN 125000 5 gasoline audi yes 24/03/2016 00:00 0 66954 07/04/2016 01:46
2 14/03/2016 12:52 9800 suv 2004 auto 163 grand 125000 8 gasoline jeep NaN 14/03/2016 00:00 0 90480 05/04/2016 12:47
3 17/03/2016 16:54 1500 small 2001 manual 75 golf 150000 6 petrol volkswagen no 17/03/2016 00:00 0 91074 17/03/2016 17:40
4 31/03/2016 17:25 3600 small 2008 manual 69 fabia 90000 7 gasoline skoda no 31/03/2016 00:00 0 60437 06/04/2016 10:17
In [6]:
# summary of columns
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 354369 entries, 0 to 354368
Data columns (total 16 columns):
 #   Column             Non-Null Count   Dtype 
---  ------             --------------   ----- 
 0   DateCrawled        354369 non-null  object
 1   Price              354369 non-null  int64 
 2   VehicleType        316879 non-null  object
 3   RegistrationYear   354369 non-null  int64 
 4   Gearbox            334536 non-null  object
 5   Power              354369 non-null  int64 
 6   Model              334664 non-null  object
 7   Mileage            354369 non-null  int64 
 8   RegistrationMonth  354369 non-null  int64 
 9   FuelType           321474 non-null  object
 10  Brand              354369 non-null  object
 11  NotRepaired        283215 non-null  object
 12  DateCreated        354369 non-null  object
 13  NumberOfPictures   354369 non-null  int64 
 14  PostalCode         354369 non-null  int64 
 15  LastSeen           354369 non-null  object
dtypes: int64(7), object(9)
memory usage: 43.3+ MB
In [7]:
# loking for missing values
df.isna().sum()
Out[7]:
DateCrawled              0
Price                    0
VehicleType          37490
RegistrationYear         0
Gearbox              19833
Power                    0
Model                19705
Mileage                  0
RegistrationMonth        0
FuelType             32895
Brand                    0
NotRepaired          71154
DateCreated              0
NumberOfPictures         0
PostalCode               0
LastSeen                 0
dtype: int64
In [8]:
# change column names
df.columns =['date_crawled', 'price', 'vehicle_type', 'registration_year', 'gearbox', 'power', 'model', 'mileage', 'registration_month', 'fuel_type', 'brand', 'not_repaired', 'date_created', 'number_of_pictures', 'postal_code', 'last_seen']
In [9]:
# Percentage missing 
df.not_repaired.isna().sum() / len(df) * 100
Out[9]:
20.079070121822166
In [10]:
# Percentage missing 
df.vehicle_type.isna().sum() / len(df) * 100
Out[10]:
10.57936783409384
In [11]:
# Percentage missing 
df.gearbox.isna().sum() / len(df) * 100
Out[11]:
5.596708515699737
In [12]:
# Percentage missing 
df.fuel_type.isna().sum() / len(df) * 100
Out[12]:
9.282696849893755
In [13]:
# changing datatypes to datetimes
df.date_crawled = pd.to_datetime(df.date_crawled)
In [14]:
# check proper implementation
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 354369 entries, 0 to 354368
Data columns (total 16 columns):
 #   Column              Non-Null Count   Dtype         
---  ------              --------------   -----         
 0   date_crawled        354369 non-null  datetime64[ns]
 1   price               354369 non-null  int64         
 2   vehicle_type        316879 non-null  object        
 3   registration_year   354369 non-null  int64         
 4   gearbox             334536 non-null  object        
 5   power               354369 non-null  int64         
 6   model               334664 non-null  object        
 7   mileage             354369 non-null  int64         
 8   registration_month  354369 non-null  int64         
 9   fuel_type           321474 non-null  object        
 10  brand               354369 non-null  object        
 11  not_repaired        283215 non-null  object        
 12  date_created        354369 non-null  object        
 13  number_of_pictures  354369 non-null  int64         
 14  postal_code         354369 non-null  int64         
 15  last_seen           354369 non-null  object        
dtypes: datetime64[ns](1), int64(7), object(8)
memory usage: 43.3+ MB
In [15]:
# extracting data from datetime columns
df['year_crawled'] = df.date_crawled.dt.year
df['month_crawled'] = df.date_crawled.dt.month
df['day_crawled'] = df.date_crawled.dt.day
df['time_crawled'] = df.date_crawled.dt.hour
In [16]:
# drop unnecessary columns
df.drop(['date_crawled', 'date_created', 'last_seen'], axis=1, inplace=True)
In [17]:
# visual of data
df.head()
Out[17]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired number_of_pictures postal_code year_crawled month_crawled day_crawled time_crawled
0 480 NaN 1993 manual 0 golf 150000 0 petrol volkswagen NaN 0 70435 2016 3 24 11
1 18300 coupe 2011 manual 190 NaN 125000 5 gasoline audi yes 0 66954 2016 3 24 10
2 9800 suv 2004 auto 163 grand 125000 8 gasoline jeep NaN 0 90480 2016 3 14 12
3 1500 small 2001 manual 75 golf 150000 6 petrol volkswagen no 0 91074 2016 3 17 16
4 3600 small 2008 manual 69 fabia 90000 7 gasoline skoda no 0 60437 2016 3 31 17
In [18]:
# look at the columns names
df.columns
Out[18]:
Index(['price', 'vehicle_type', 'registration_year', 'gearbox', 'power',
       'model', 'mileage', 'registration_month', 'fuel_type', 'brand',
       'not_repaired', 'number_of_pictures', 'postal_code', 'year_crawled',
       'month_crawled', 'day_crawled', 'time_crawled'],
      dtype='object')

Conclusions¶

We see many missing values in categorical columns, which would need to be filled for some of our machine learning models. Vehicle type may be imputed based on the mode of each model and brand of the car. The columns gearbox, model, fuel type, and not repaired will be filled based on the mode of the respective columns. Issues may arise where rows have missing values in multiple columns. Such rows will just be dropped.

Data Analysis¶

In [19]:
# correlation of categories
df.corr()
Out[19]:
price registration_year power mileage registration_month number_of_pictures postal_code year_crawled month_crawled day_crawled time_crawled
price 1.000000 0.026916 0.158872 -0.333199 0.110581 NaN 0.076055 NaN -0.007529 -0.003374 0.017560
registration_year 0.026916 1.000000 -0.000828 -0.053447 -0.011619 NaN -0.003459 NaN -0.001598 0.001513 -0.001617
power 0.158872 -0.000828 1.000000 0.024002 0.043380 NaN 0.021665 NaN 0.000750 -0.001815 -0.000617
mileage -0.333199 -0.053447 0.024002 1.000000 0.009571 NaN -0.007698 NaN 0.000245 0.001013 -0.011518
registration_month 0.110581 -0.011619 0.043380 0.009571 1.000000 NaN 0.013995 NaN 0.003955 -0.002674 0.012123
number_of_pictures NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
postal_code 0.076055 -0.003459 0.021665 -0.007698 0.013995 NaN 1.000000 NaN -0.010531 0.004484 0.016590
year_crawled NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
month_crawled -0.007529 -0.001598 0.000750 0.000245 0.003955 NaN -0.010531 NaN 1.000000 -0.572049 -0.000910
day_crawled -0.003374 0.001513 -0.001815 0.001013 -0.002674 NaN 0.004484 NaN -0.572049 1.000000 0.006911
time_crawled 0.017560 -0.001617 -0.000617 -0.011518 0.012123 NaN 0.016590 NaN -0.000910 0.006911 1.000000
In [20]:
# summary statistics of data
df.describe()
Out[20]:
price registration_year power mileage registration_month number_of_pictures postal_code year_crawled month_crawled day_crawled time_crawled
count 354369.000000 354369.000000 354369.000000 354369.000000 354369.000000 354369.0 354369.000000 354369.0 354369.000000 354369.000000 354369.000000
mean 4416.656776 2004.234448 110.094337 128211.172535 5.714645 0.0 50508.689087 2016.0 4.406150 14.609021 14.909665
std 4514.158514 90.227958 189.850405 37905.341530 3.726421 0.0 25783.096248 0.0 2.804298 10.190826 5.422938
min 0.000000 1000.000000 0.000000 5000.000000 0.000000 0.0 1067.000000 2016.0 1.000000 3.000000 0.000000
25% 1050.000000 1999.000000 69.000000 125000.000000 3.000000 0.0 30165.000000 2016.0 3.000000 4.000000 12.000000
50% 2700.000000 2003.000000 105.000000 150000.000000 6.000000 0.0 49413.000000 2016.0 3.000000 16.000000 16.000000
75% 6400.000000 2008.000000 143.000000 150000.000000 9.000000 0.0 71083.000000 2016.0 5.000000 24.000000 19.000000
max 20000.000000 9999.000000 20000.000000 150000.000000 12.000000 0.0 99998.000000 2016.0 12.000000 31.000000 23.000000
In [21]:
# number of picture values
df.number_of_pictures.value_counts(dropna=False)
Out[21]:
0    354369
Name: number_of_pictures, dtype: int64
In [22]:
# drop unnecessary column
df.drop('number_of_pictures', axis=1, inplace=True)

Unnecessary column, so we will drop it.

In [23]:
# dropping rows where vehicle type and model are missing
mask = (df['vehicle_type'].isnull() & df['model'].isnull())
df.drop(df[mask].index, inplace=True)
In [24]:
# Check for proper implementation
(df['vehicle_type'].isnull() & df['model'].isnull()).sum()
Out[24]:
0

Dropped rows where vehicle type and model values were both missing.

In [25]:
# petrol and gasoline are the same
df.fuel_type = df.fuel_type.replace('gasoline', 'petrol')

Gasoline and petrol are the same thing. Gasoline is just the American way of saying petrol.

In [26]:
# number of cars with registration years below 1960
df.query("registration_year < 1960")
Out[26]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
15 450 small 1910 NaN 0 ka 5000 0 petrol ford NaN 24148 2016 11 3 21
1928 7000 suv 1945 manual 48 other 150000 2 petrol volkswagen no 58135 2016 3 25 15
2273 1800 convertible 1925 NaN 0 NaN 5000 1 NaN sonstige_autos no 79288 2016 3 15 21
3333 10500 sedan 1955 manual 30 other 60000 0 petrol ford NaN 53498 2016 3 15 21
6629 0 small 1910 NaN 0 NaN 5000 1 other sonstige_autos NaN 93105 2016 2 4 13
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
351183 17500 sedan 1954 manual 52 other 20000 7 petrol citroen no 55270 2016 3 27 20
351299 5500 bus 1956 manual 37 NaN 60000 4 petrol sonstige_autos no 1900 2016 9 3 21
351682 11500 NaN 1800 NaN 16 other 5000 6 petrol fiat NaN 16515 2016 12 3 0
353531 6000 sedan 1937 manual 38 other 5000 0 petrol mercedes_benz NaN 23936 2016 3 16 21
354316 3300 coupe 1957 manual 40 other 100000 11 petrol trabant no 10317 2016 7 3 19

339 rows × 16 columns

In [27]:
# filtering outliers of dataset
df = df.query("(registration_year > 1960) and (registration_year <= 2019)")

We will remove outliers from the dataset based on registration years. The first automobile in europe was made in 1897, yet, it is reasonable to assume that cars from that period would not be on the Rusty Bargain app. Furthermore, the number of cars in the dataset from before 1960 is limited, so we will not be losing much data. Also, registration years above 2019 would make those cars registered decades in the future. Removing these outliers will make our models perform better.

In [28]:
# looking for outliers
df.registration_year.value_counts()
Out[28]:
2000    23177
1999    22717
2005    21566
2001    20115
2006    19830
2003    19754
2004    19598
2002    19087
1998    17880
2007    17142
2008    15358
1997    14649
2009    14460
1996    10823
2010    10723
2011     9884
1995     9170
2017     9015
2016     8318
2012     7166
1994     4840
2013     4114
1993     3434
2018     3386
1992     3001
1991     2761
2014     2519
1990     2501
1989     1317
2015     1192
1988      941
1985      684
1987      666
1986      537
1980      482
1983      430
1984      399
1982      316
1978      285
1979      283
1970      272
1981      263
1972      213
1976      175
1977      173
1973      173
1974      168
1971      167
1975      143
1969      118
1966      111
1968      108
1967       99
1965       82
1964       64
1963       57
1962       46
1961       41
2019       20
Name: registration_year, dtype: int64
In [29]:
# boxplot of registration
px.box(df.registration_year)
registration_year1960197019801990200020102020
variablevalue
plotly-logomark
In [30]:
# values of power
df.power.value_counts()
Out[30]:
0       36270
75      23721
60      15687
150     14443
101     13158
        ...  
2789        1
519         1
2461        1
6006        1
1241        1
Name: power, Length: 706, dtype: int64
In [31]:
# looking at outliers of power
px.box(df.power)
power05k10k15k20k
variablevalue
plotly-logomark
In [32]:
# filtering outliers of dataset
df = df.query("(power > 20) and (power < 500)")

We also remove outliers based on power. Power values of zero do not make sense, as even electric vehicles have power. Alternatively, power above 500 hp is generally reserved for super and hypercars.

In [33]:
# filter for categories
categories = ['price', 'vehicle_type', 'registration_year', 'gearbox', 'power',
       'model', 'mileage', 'registration_month', 'fuel_type', 'brand',
       'not_repaired','year_crawled', 'month_crawled', 'day_crawled', 'time_crawled']
In [34]:
# loop for displaying histograms of categories 
for i in categories:
    px.histogram(df[i], title='Distribution of ' + str.upper(i).replace('_', ' '), labels={'value': i}, template='seaborn').show()
05k10k15k20k02000400060008000
variablepriceDistribution of PRICEpricecount
plotly-logomark
coupesuvsmallsedanconvertiblebuswagonother010k20k30k40k50k60k70k80k
variablevehicle_typeDistribution of VEHICLE TYPEvehicle_typecount
plotly-logomark
1970198019902000201005k10k15k20k
variableregistration_yearDistribution of REGISTRATION YEARregistration_yearcount
plotly-logomark
manualauto050k100k150k200k250k
variablegearboxDistribution of GEARBOXgearboxcount
plotly-logomark
5010015020025030035040045005k10k15k20k
variablepowerDistribution of POWERpowercount
plotly-logomark
grandfabia2_reihec_maxpassatpoloa_klasse5erarosacivicpuntoclioone1era8astrac_klassevito156foresterscenica1a4a6omega7ercorsa147sportageibizaeosgetzalmeraa3r19mondeocoltvectratiguan6_reiheupkangooyetizafirarx_reihemodusmatizriologancuorea2c3s_klassesharanroomsterkaeferleon4_reihelagunaclkespace159jukev40accordlanosversoravpicantosuperbalhambraroadstergalant90signumagilav50c_reihem_klassec5seicentocarnival911forfourcl300ccr_reihekugaducatox_trailalteacalibrabravoc1jimnywranglerlancerc2freelanderq7antaracherokeedefendercx_reihepajeronivanubirag_klasserange_roverq3glkcromadoblo9000v60200b_maxmoveq5materiaterioskalinai3amarokserie_305k10k15k20k25k
variablemodelDistribution of MODELmodelcount
plotly-logomark
20k40k60k80k100k120k140k050k100k150k200k
variablemileageDistribution of MILEAGEmileagecount
plotly-logomark
02468101205k10k15k20k25k30k
variableregistration_monthDistribution of REGISTRATION MONTHregistration_monthcount
plotly-logomark
petrollpgotherhybridcngelectric050k100k150k200k250k
variablefuel_typeDistribution of FUEL TYPEfuel_typecount
plotly-logomark
audijeepvolkswagenskodabmwpeugeotfordmazdanissanrenaultmercedes_benzopelseatcitroenhondafiatminismarthyundaisonstige_autosalfa_romeosubaruvolvomitsubishikiasuzukilanciatoyotachevroletdaciadaihatsusaabchryslerjaguardaewooroverporscheland_rovertrabantlada010k20k30k40k50k60k70k
variablebrandDistribution of BRANDbrandcount
plotly-logomark
yesno050k100k150k200k
variablenot_repairedDistribution of NOT REPAIREDnot_repairedcount
plotly-logomark
2,015.62,015.820162,016.22,016.4050k100k150k200k250k300k
variableyear_crawledDistribution of YEAR CRAWLEDyear_crawledcount
plotly-logomark
24681012050k100k150k200k
variablemonth_crawledDistribution of MONTH CRAWLEDmonth_crawledcount
plotly-logomark
51015202530010k20k30k40k50k60k70k80k
variableday_crawledDistribution of DAY CRAWLEDday_crawledcount
plotly-logomark
0510152005k10k15k20k
variabletime_crawledDistribution of TIME CRAWLEDtime_crawledcount
plotly-logomark

Distribution of price shows some vehicles that have a value of 0. As this is historical data, it appears as if Rusty Bargain has determined the market values of these vehicles to be 0. Distribution of vehicle type is distributed around small, sedan, and wagon. The distribution of registration years is now left skewed, with the median around the mid 2000's. There are two options for gearbox: manual and auto. Most of the cars are manual transmission. Power is right skewed, with most values falling between 50 and 200 hp. Mileage is left skewed, with most cars having 150K kilometers. Fuel type is mostly petrol, and most cars are not repaired.

In [35]:
# missing values
df.isna().sum()
Out[35]:
price                     0
vehicle_type          19582
registration_year         0
gearbox                5989
power                     0
model                  9950
mileage                   0
registration_month        0
fuel_type             19290
brand                     0
not_repaired          47864
postal_code               0
year_crawled              0
month_crawled             0
day_crawled               0
time_crawled              0
dtype: int64
In [36]:
# fill not repaired missing values with no
df.not_repaired.fillna('no', inplace=True)

Assumption made that missing values in not repaired column, are not repaired.

In [37]:
# check proper implementation
df.not_repaired.value_counts(dropna=False)
Out[37]:
no     279108
yes     30780
Name: not_repaired, dtype: int64
In [38]:
# looking at missing gearbox values
df.gearbox.value_counts(dropna=False)
Out[38]:
manual    243171
auto       60728
NaN         5989
Name: gearbox, dtype: int64

Assumption made that missing values in the gearbox column are manual transmissions.

In [39]:
# looking at range of registration year
df.registration_year.value_counts().sort_values(axis=0, ascending=False)
Out[39]:
1999    20122
2000    19072
2005    18982
2006    18649
2001    18243
2003    18128
2004    18000
2002    17462
2007    16174
1998    15720
2008    14592
2009    13874
1997    12948
2010    10381
2011     9594
1996     9505
1995     7482
2012     6935
2017     6809
2016     5911
1994     4281
2013     3986
1993     3034
1992     2680
2018     2593
1991     2435
2014     2403
1990     1913
1989     1136
2015     1075
1988      826
1987      566
1985      490
1986      434
1983      360
1984      343
1982      277
1980      255
1979      236
1978      226
1981      225
1972      160
1977      139
1973      129
1976      129
1974      126
1970      120
1971      120
1975      104
1969       89
1968       85
1966       80
1967       62
1965       48
1964       41
1963       38
1962       26
1961       25
2019       10
Name: registration_year, dtype: int64
In [40]:
# query for price is 0
df.query("price == 0")
Out[40]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
7 0 sedan 1980 manual 50 other 40000 7 petrol volkswagen no 19348 2016 3 21 18
152 0 bus 2004 manual 101 meriva 150000 10 lpg opel yes 27432 2016 11 3 18
231 0 wagon 2001 manual 115 mondeo 150000 0 NaN ford no 57627 2016 3 16 22
466 0 NaN 2016 auto 197 3er 150000 12 petrol bmw no 99867 2016 3 24 10
563 0 sedan 1998 auto 230 NaN 150000 10 petrol saab no 71691 2016 3 21 12
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
354175 0 NaN 1995 manual 45 polo 150000 0 petrol volkswagen no 2625 2016 3 20 17
354205 0 NaN 2000 manual 65 corsa 150000 0 NaN opel yes 23758 2016 9 3 15
354238 0 small 2002 manual 60 fiesta 150000 3 petrol ford no 33659 2016 3 20 14
354248 0 small 1999 manual 53 swift 150000 3 petrol suzuki no 42329 2016 3 24 13
354277 0 small 1999 manual 37 arosa 150000 7 petrol seat yes 22559 2016 10 3 22

6539 rows × 16 columns

Conclusions¶

We've used EDA to analyze the distributions and outliers in the data. We have removed many of the values that did not make intuitive sense, in an attempt to improve the accuracy of the modeling.

Model Preparation¶

In [41]:
# making copy of original dataset
df_copy = df.copy()
In [42]:
# Select only the object columns from the DataFrame
object_columns = df_copy.select_dtypes(include='object')
In [43]:
# fill missing obkect columns with tring NaN
object_columns.fillna('NaN', inplace=True)
In [44]:
# label encoding categorical columns
encoder = OrdinalEncoder()
ordinal = pd.DataFrame(encoder.fit_transform(object_columns), columns=object_columns.columns)
In [45]:
# checking implementation
ordinal.head()
Out[45]:
vehicle_type gearbox model fuel_type brand not_repaired
0 3.0 2.0 26.0 6.0 1.0 1.0
1 7.0 1.0 118.0 6.0 14.0 0.0
2 6.0 2.0 117.0 6.0 38.0 0.0
3 6.0 2.0 102.0 6.0 31.0 0.0
4 5.0 2.0 11.0 6.0 2.0 1.0
In [46]:
# check the ordinal index parameters
ordinal.index
Out[46]:
RangeIndex(start=0, stop=309888, step=1)
In [47]:
# start index at 1
ordinal.index = ordinal.index + 1
In [48]:
# check implementation
ordinal
Out[48]:
vehicle_type gearbox model fuel_type brand not_repaired
1 3.0 2.0 26.0 6.0 1.0 1.0
2 7.0 1.0 118.0 6.0 14.0 0.0
3 6.0 2.0 117.0 6.0 38.0 0.0
4 6.0 2.0 102.0 6.0 31.0 0.0
5 5.0 2.0 11.0 6.0 2.0 1.0
... ... ... ... ... ... ...
309884 0.0 1.0 4.0 0.0 0.0 0.0
309885 5.0 2.0 141.0 6.0 30.0 1.0
309886 2.0 1.0 107.0 6.0 32.0 0.0
309887 1.0 2.0 224.0 6.0 38.0 0.0
309888 8.0 2.0 117.0 6.0 38.0 0.0

309888 rows × 6 columns

In [49]:
# checking index
ordinal.index
Out[49]:
RangeIndex(start=1, stop=309889, step=1)
In [50]:
# new ordinal shape
ordinal.shape
Out[50]:
(309888, 6)
In [51]:
# compare to original df shape
df.shape
Out[51]:
(309888, 16)
In [52]:
# Replace the original object columns in the DataFrame with the encoded columns
df_copy[object_columns.columns] = ordinal
In [53]:
# for comparisons
df.head()
Out[53]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
1 18300 coupe 2011 manual 190 NaN 125000 5 petrol audi yes 66954 2016 3 24 10
2 9800 suv 2004 auto 163 grand 125000 8 petrol jeep no 90480 2016 3 14 12
3 1500 small 2001 manual 75 golf 150000 6 petrol volkswagen no 91074 2016 3 17 16
4 3600 small 2008 manual 69 fabia 90000 7 petrol skoda no 60437 2016 3 31 17
5 650 sedan 1995 manual 102 3er 150000 10 petrol bmw yes 33775 2016 4 4 17
In [54]:
# check full implementation of label encoding
df_copy.head()
Out[54]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
1 18300 3.0 2011 2.0 190 26.0 125000 5 6.0 1.0 1.0 66954 2016 3 24 10
2 9800 7.0 2004 1.0 163 118.0 125000 8 6.0 14.0 0.0 90480 2016 3 14 12
3 1500 6.0 2001 2.0 75 117.0 150000 6 6.0 38.0 0.0 91074 2016 3 17 16
4 3600 6.0 2008 2.0 69 102.0 90000 7 6.0 31.0 0.0 60437 2016 3 31 17
5 650 5.0 1995 2.0 102 11.0 150000 10 6.0 2.0 1.0 33775 2016 4 4 17

We use ordinal encoding to change the categorical values into numerical labels.

In [55]:
# using iterative imputer
imp = IterativeImputer()
data_imp = imp.fit_transform(df_copy)
In [56]:
# making a new dataframe for encoded data, rounding to nearest integer
df1 = pd.DataFrame(data_imp, columns=df.columns)
df1 = df1.round({'vehicle_type':0, 'gearbox':0, 'model':0, 'brand':0, 'not_repaired':0, 'fuel_type':0})
In [57]:
# visual of new dataset
df1.head()
Out[57]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
0 18300.0 3.0 2011.0 2.0 190.0 26.0 125000.0 5.0 6.0 1.0 1.0 66954.0 2016.0 3.0 24.0 10.0
1 9800.0 7.0 2004.0 1.0 163.0 118.0 125000.0 8.0 6.0 14.0 0.0 90480.0 2016.0 3.0 14.0 12.0
2 1500.0 6.0 2001.0 2.0 75.0 117.0 150000.0 6.0 6.0 38.0 0.0 91074.0 2016.0 3.0 17.0 16.0
3 3600.0 6.0 2008.0 2.0 69.0 102.0 90000.0 7.0 6.0 31.0 0.0 60437.0 2016.0 3.0 31.0 17.0
4 650.0 5.0 1995.0 2.0 102.0 11.0 150000.0 10.0 6.0 2.0 1.0 33775.0 2016.0 4.0 4.0 17.0

Next, we use iterative imputer to work on missing values. We chose iterative imputer instead of a simple imputer because we did not want the missing values to be replaced by calculations of just a single column. Iterative imputer fills in missing values based on multiple columns. This should create a better fill value. Then, we round the calculated missing values to the nearest integer, to match the rest of the data. Overall, we are left with roughly 300,000 data points to use for modelling.

In [58]:
# checking missing values
df1.isna().sum()
Out[58]:
price                 0
vehicle_type          0
registration_year     0
gearbox               0
power                 0
model                 0
mileage               0
registration_month    0
fuel_type             0
brand                 0
not_repaired          0
postal_code           0
year_crawled          0
month_crawled         0
day_crawled           0
time_crawled          0
dtype: int64

New, imputed dataframe does not contain missing values

In [59]:
# looking at vehicle type values
df.vehicle_type.value_counts(sort=True, dropna=False)
Out[59]:
sedan          84603
small          71226
wagon          60397
bus            26483
NaN            19582
convertible    18983
coupe          15018
suv            11064
other           2532
Name: vehicle_type, dtype: int64
In [60]:
# checking vehicle type value counts
df1.vehicle_type.value_counts(sort=True, dropna=False)
Out[60]:
5.0    112928
6.0     62216
8.0     52879
1.0     23181
0.0     17102
2.0     16579
3.0     13117
7.0      9674
4.0      2212
Name: vehicle_type, dtype: int64

We compare the number of different values in the original and imputed dataframes. Overall, there is a similar amount of types, and the counts see some changes, with the most popular type increasing the most.

In [61]:
# visual of original data
df.head(20)
Out[61]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
1 18300 coupe 2011 manual 190 NaN 125000 5 petrol audi yes 66954 2016 3 24 10
2 9800 suv 2004 auto 163 grand 125000 8 petrol jeep no 90480 2016 3 14 12
3 1500 small 2001 manual 75 golf 150000 6 petrol volkswagen no 91074 2016 3 17 16
4 3600 small 2008 manual 69 fabia 90000 7 petrol skoda no 60437 2016 3 31 17
5 650 sedan 1995 manual 102 3er 150000 10 petrol bmw yes 33775 2016 4 4 17
6 2200 convertible 2004 manual 109 2_reihe 150000 8 petrol peugeot no 67112 2016 1 4 20
7 0 sedan 1980 manual 50 other 40000 7 petrol volkswagen no 19348 2016 3 21 18
8 14500 bus 2014 manual 125 c_max 30000 8 petrol ford no 94505 2016 4 4 23
9 999 small 1998 manual 101 golf 150000 0 NaN volkswagen no 27472 2016 3 17 10
10 2000 sedan 2004 manual 105 3_reihe 150000 12 petrol mazda no 96224 2016 3 26 19
11 2799 wagon 2005 manual 140 passat 150000 12 petrol volkswagen yes 57290 2016 7 4 10
12 999 wagon 1995 manual 115 passat 150000 11 petrol volkswagen no 37269 2016 3 15 22
13 2500 wagon 2004 manual 131 passat 150000 2 NaN volkswagen no 90762 2016 3 21 21
14 17999 suv 2011 manual 190 navara 70000 3 petrol nissan no 4177 2016 3 21 12
16 300 NaN 2016 NaN 60 polo 150000 0 petrol volkswagen no 38871 2016 1 4 12
17 1750 small 2004 auto 75 twingo 150000 2 petrol renault no 65599 2016 3 20 10
18 7550 bus 2007 manual 136 c_max 150000 6 petrol ford no 88361 2016 3 23 15
19 1850 bus 2004 manual 102 a_klasse 150000 1 petrol mercedes_benz no 49565 2016 1 4 22
20 10400 coupe 2009 manual 160 scirocco 100000 4 petrol volkswagen no 75365 2016 1 4 19
21 3699 sedan 2002 auto 231 5er 150000 7 petrol bmw no 68309 2016 3 27 11
In [62]:
# small cars
df.query("vehicle_type == 'small'")
Out[62]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
3 1500 small 2001 manual 75 golf 150000 6 petrol volkswagen no 91074 2016 3 17 16
4 3600 small 2008 manual 69 fabia 90000 7 petrol skoda no 60437 2016 3 31 17
9 999 small 1998 manual 101 golf 150000 0 NaN volkswagen no 27472 2016 3 17 10
17 1750 small 2004 auto 75 twingo 150000 2 petrol renault no 65599 2016 3 20 10
23 450 small 1997 manual 50 arosa 150000 5 petrol seat no 9526 2016 12 3 19
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
354326 1300 small 1999 manual 75 2_reihe 125000 0 NaN peugeot no 35102 2016 3 31 19
354329 350 small 1996 NaN 65 punto 150000 0 NaN fiat no 25436 2016 3 30 20
354342 600 small 1998 manual 54 corsa 150000 1 petrol opel no 24850 2016 8 3 13
354345 1700 small 1999 manual 68 justy 70000 11 petrol subaru no 90439 2016 1 4 1
354358 1490 small 1998 manual 50 lupo 150000 9 petrol volkswagen no 48653 2016 4 4 9

71226 rows × 16 columns

In [63]:
# visual of encoded data
df1.head()
Out[63]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
0 18300.0 3.0 2011.0 2.0 190.0 26.0 125000.0 5.0 6.0 1.0 1.0 66954.0 2016.0 3.0 24.0 10.0
1 9800.0 7.0 2004.0 1.0 163.0 118.0 125000.0 8.0 6.0 14.0 0.0 90480.0 2016.0 3.0 14.0 12.0
2 1500.0 6.0 2001.0 2.0 75.0 117.0 150000.0 6.0 6.0 38.0 0.0 91074.0 2016.0 3.0 17.0 16.0
3 3600.0 6.0 2008.0 2.0 69.0 102.0 90000.0 7.0 6.0 31.0 0.0 60437.0 2016.0 3.0 31.0 17.0
4 650.0 5.0 1995.0 2.0 102.0 11.0 150000.0 10.0 6.0 2.0 1.0 33775.0 2016.0 4.0 4.0 17.0
In [64]:
# type 4 cars
df1.query('vehicle_type == 4')
Out[64]:
price vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code year_crawled month_crawled day_crawled time_crawled
31 3500.0 4.0 2004.0 2.0 122.0 130.0 150000.0 11.0 0.0 24.0 0.0 67071.0 2016.0 3.0 17.0 18.0
32 350.0 4.0 2016.0 2.0 75.0 167.0 150000.0 4.0 6.0 38.0 0.0 19386.0 2016.0 8.0 3.0 7.0
61 1280.0 4.0 1992.0 2.0 109.0 237.0 150000.0 9.0 6.0 20.0 0.0 86163.0 2016.0 3.0 15.0 9.0
315 16500.0 4.0 1983.0 2.0 218.0 26.0 150000.0 10.0 6.0 0.0 0.0 42719.0 2016.0 3.0 30.0 13.0
331 2950.0 4.0 2005.0 2.0 131.0 167.0 150000.0 6.0 6.0 5.0 0.0 91781.0 2016.0 12.0 3.0 12.0
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
270625 8990.0 4.0 2007.0 2.0 150.0 214.0 150000.0 6.0 6.0 9.0 0.0 49163.0 2016.0 10.0 3.0 17.0
270664 1000.0 4.0 2016.0 2.0 175.0 80.0 150000.0 0.0 6.0 24.0 1.0 59174.0 2016.0 9.0 3.0 1.0
270694 800.0 4.0 2016.0 2.0 95.0 224.0 150000.0 1.0 6.0 38.0 0.0 72070.0 2016.0 7.0 3.0 11.0
270729 4500.0 4.0 1997.0 2.0 150.0 237.0 150000.0 4.0 6.0 20.0 0.0 22609.0 2016.0 4.0 4.0 17.0
270965 200.0 4.0 1997.0 2.0 102.0 10.0 150000.0 10.0 6.0 25.0 0.0 36154.0 2016.0 3.0 14.0 23.0

2212 rows × 16 columns

In [65]:
df1.columns
Out[65]:
Index(['price', 'vehicle_type', 'registration_year', 'gearbox', 'power',
       'model', 'mileage', 'registration_month', 'fuel_type', 'brand',
       'not_repaired', 'postal_code', 'year_crawled', 'month_crawled',
       'day_crawled', 'time_crawled'],
      dtype='object')
In [66]:
# maybe look to eliminate more useless columns from features
features = df1.drop(['price','year_crawled', 'month_crawled',
       'day_crawled', 'time_crawled'], axis=1)
target = df1['price']
In [67]:
# train test split
features_train, features_test, target_train, target_test = train_test_split(
    features, target, test_size=0.25, random_state=19)
In [68]:
# train validate split
features_train, features_valid, target_train, target_valid = train_test_split(
    features_train, target_train, test_size=0.2, random_state=19)
In [69]:
# Scaling numerical columns
numeric = ['registration_year', 'power', 'mileage', 'postal_code']

scaler = StandardScaler()
scaler.fit(features_train[numeric])

features_train[numeric] = scaler.transform(features_train[numeric])
features_test[numeric] = scaler.transform(features_test[numeric])
In [70]:
# Creating new features to test prediction times
new_features = pd.DataFrame(
    [
        [4, 1999, 1, 75, 108, 150000, 3, 5, 38, 0, 10115],
        [5, 2008, 1, 150, 166, 150000, 9, 5, 31, 0, 71672]
    ],
    columns=features.columns
)
In [71]:
# for regressions
new_features
Out[71]:
vehicle_type registration_year gearbox power model mileage registration_month fuel_type brand not_repaired postal_code
0 4 1999 1 75 108 150000 3 5 38 0 10115
1 5 2008 1 150 166 150000 9 5 31 0 71672

Conclusions¶

We encoded the categorical values into integers, and then filled missing values via imputation. The dataset no longer contains missing values, a condition that is required for some of our models to work.

Modeling¶

Regressions¶

Decision Tree¶

In [72]:
# %%timeit
# decision tree
best_model = None
best_result = 10000
best_depth = 0
for depth in range(1, 15): # choose hyperparameter range
    dec_tr = DecisionTreeRegressor(random_state=19, max_depth=depth)
    dec_tr.fit(features_train, target_train) # train model on training set
    predictions_valid_dt = dec_tr.predict(features_valid) # get model predictions on validation set
    result = mse(target_valid, predictions_valid_dt) ** 0.5
    if result < best_result:
        best_model = dec_tr
        best_result = result
        best_depth = depth

print(f"RMSE of the best model on the validation set (max_depth = {best_depth}): {best_result}")

RMSE of the best model on the validation set (max_depth = 5): 4689.009293240116
In [73]:
# %%timeit
# Decision tree prediction
dec_tr.predict(new_features)
Out[73]:
array([16500., 17990.])

Random Forest¶

In [74]:
# %%timeit
# random forest
best_model = None
best_result = 10000
best_est = 0
best_depth = 0
for est in range(40, 51, 5):
    for depth in range (10, 15, 5):
        rf = RandomForestRegressor(random_state=19, n_estimators=est, max_depth=depth)
        rf.fit(features_train, target_train) # train model on training set
        predictions_valid_rf = rf.predict(features_valid) # get model predictions on validation set
        result = mse(target_valid, predictions_valid_rf) ** 0.5 # calculate RMSE on validation set
        if result < best_result:
            best_model = rf
            best_result = result
            best_est = est
            best_depth = depth

print("RMSE of the best model on the validation set:", best_result, "n_estimators:", best_est, "best_depth:", depth)

RMSE of the best model on the validation set: 8585.188109611085 n_estimators: 45 best_depth: 10
In [75]:
# %%timeit
# random forest prediction
rf.predict(new_features)
Out[75]:
array([12314.99208565, 13272.2579312 ])

Linear Regression¶

In [76]:
# %%timeit
# Linear regression
lr = LinearRegression() # initialize model constructor
lr.fit(features_train, target_train) # train model on training set
predictions_valid_lr = lr.predict(features_valid) # get model predictions on validation set

result = mse(target_valid, predictions_valid_lr) ** 0.5 # calculate RMSE on validation set
print("RMSE of the linear regression model on the validation set:", result)

RMSE of the linear regression model on the validation set: 206533854.8212574
In [77]:
#%%timeit
# linear regression prediction
lr.predict(new_features)
Out[77]:
array([-2.37722899e+08, -2.30083006e+08])

Comparing Regressions¶

In [78]:
# creating table of regression metrics
data = {'decision tree': (2229.78, 10.5, 1.53, 0.00234, 5.18 * 10 ** -4),  'random forest': (2172.63, 77, 10.7, 0.00426, 1.58 * 10 ** -4), 'linear regression': (3177.86, .0928, .0691, 0.00125, 3.01 * 10 ** -4)}
regr_scores = pd.DataFrame(data=data, columns=data.keys(), index=['RMSE', 'train_time', 'train_std', 'pred_time', 'pred_std'])
regr_scores = regr_scores.T
regr_scores
Out[78]:
RMSE train_time train_std pred_time pred_std
decision tree 2229.78 10.5000 1.5300 0.00234 0.000518
random forest 2172.63 77.0000 10.7000 0.00426 0.000158
linear regression 3177.86 0.0928 0.0691 0.00125 0.000301
In [79]:
# regression RMSE
px.scatter(regr_scores.RMSE, color_discrete_sequence=[['orange', 'green', 'indigo']], labels={'index': 'Model', 'value': 'RMSE Score'}, title='Model Score Comparison', size=regr_scores.RMSE, log_y=True)
decision treerandom forestlinear regression2100220023002400250026002700280029003000310032003300
variableRMSEModel Score ComparisonModelRMSE Score
plotly-logomark
In [80]:
# comparing train and predict times
fig = go.Figure()
fig.add_trace(go.Bar(
    name='Train',
    x=['decision tree', 'random forest', 'linear regression'], y=regr_scores.train_time,
    error_y=dict(type='data', array=[1.5, 10.7, .0691])
))
fig.add_trace(go.Bar(
    name='Predict',
    x=['decision tree', 'random forest', 'linear regression'], y=regr_scores.pred_time,
    error_y=dict(type='data', array=[.00518, .00158, .00301]
)))

fig.update_layout(barmode='group', title='Model Prediction Time Comparison', template='seaborn')
fig.show()
decision treerandom forestlinear regression020406080
TrainPredictModel Prediction Time Comparison
plotly-logomark
Conclusions¶

The model with the best RMSE score was the random forest, while the worst performing model was the linear regression. In regards to time, the random forest model took the longest to train, while the linear regression model was the quickest to train. The same trend was seen with model predictions.

Boosting¶

XG Boost¶

In [81]:
# %%timeit
# XGB

X2 = df.drop(['price','year_crawled', 'month_crawled',
       'day_crawled', 'time_crawled'], axis=1)
y2 = df['price']

lbl = LabelEncoder()
X2.vehicle_type = lbl.fit_transform(X2.vehicle_type.astype(str))
X2.gearbox = lbl.fit_transform(X2.gearbox.astype(str))
X2.model = lbl.fit_transform(X2.model.astype(str))
X2.fuel_type = lbl.fit_transform(X2.fuel_type.astype(str))
X2.brand = lbl.fit_transform(X2.brand.astype(str))
X2.not_repaired = lbl.fit_transform(X2.not_repaired.astype(str))


X2_train, X2_test, y2_train, y2_test = train_test_split(
    X2, y2, test_size=0.25, random_state=19)

X2_train, X2_valid, y2_train, y2_valid = train_test_split(
    X2_train, y2_train, test_size=0.20, random_state=19)


# Create a XGBM 
xgbr = xgb.XGBRegressor(random_state=19, eval_metric='rmse')

xgbr.fit(X2_train, y2_train)

# Make predictions on the test set
predictions_xgbr = xgbr.predict(X2_valid)

result = mse(y2_valid, predictions_xgbr) ** 0.5 # calculate RMSE on validation set
print()
print("RMSE of the xgbm model on the validation set:", result)

RMSE of the xgbm model on the validation set: 1743.2028538931468
In [82]:
%%timeit
# xg boost prediction
predictions_xgbr = xgbr.predict(new_features)

4.7 ms ± 42.4 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
In [83]:
# xg boost parameters
xgbr.get_xgb_params
Out[83]:
<bound method XGBModel.get_xgb_params of XGBRegressor(base_score=0.5, booster='gbtree', callbacks=None,
             colsample_bylevel=1, colsample_bynode=1, colsample_bytree=1,
             early_stopping_rounds=None, enable_categorical=False,
             eval_metric='rmse', feature_types=None, gamma=0, gpu_id=-1,
             grow_policy='depthwise', importance_type=None,
             interaction_constraints='', learning_rate=0.300000012, max_bin=256,
             max_cat_threshold=64, max_cat_to_onehot=4, max_delta_step=0,
             max_depth=6, max_leaves=0, min_child_weight=1, missing=nan,
             monotone_constraints='()', n_estimators=100, n_jobs=0,
             num_parallel_tree=1, predictor='auto', random_state=19, ...)>
In [84]:
# table of feature importance
xgbr_imp = [t for t in zip(features, xgbr.feature_importances_)]
xgbr_imp_df = pd.DataFrame(xgbr_imp, columns=['feature', 'varimp'])
xgbr_imp_df = xgbr_imp_df.sort_values('varimp', ascending=False)
xgbr_imp_df
Out[84]:
feature varimp
1 registration_year 0.355426
3 power 0.200501
5 mileage 0.146589
9 not_repaired 0.101944
8 brand 0.052967
0 vehicle_type 0.051839
2 gearbox 0.038437
4 model 0.021239
7 fuel_type 0.012837
6 registration_month 0.009155
10 postal_code 0.009067

Light GBM¶

In [85]:
# %%timeit
# LGBM

X = df.drop(['price','year_crawled', 'month_crawled',
       'day_crawled', 'time_crawled'], axis=1)
y = df['price']

obj_feat = list(X.loc[:, X.dtypes == 'object'].columns.values)

for feature in obj_feat:
    X[feature] = pd.Series(X[feature], dtype='category')

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.25, random_state=19)

X_train, X_valid, y_train, y_valid = train_test_split(
    X_train, y_train, test_size=0.20, random_state=19)

# Create a LightGBM dataset
lgb_train = lgb.Dataset(X_train, y_train)
lgb_valid = lgb.Dataset(X_valid, y_valid, reference=lgb_train)

# Define the parameters for the LightGBM model
params = {
    'objective': 'regression',
    'metric': 'root_mean_squared_error',
    'boosting_type': 'gbdt',
    'random_state': 19
}

# Train the LightGBM model
lgbm = lgb.train(params, lgb_train, valid_sets=lgb_valid, num_boost_round=5000, early_stopping_rounds=500)

# Make predictions on the validation set
predictions_valid_lgbm = lgbm.predict(X_valid)

result = mse(y_valid, predictions_valid_lgbm) ** 0.5 # calculate RMSE on validation set
print()
print("RMSE of the lgbm model on the validation set:", result)

C:\Users\XIX\anaconda3\lib\site-packages\lightgbm\engine.py:181: UserWarning:

'early_stopping_rounds' argument is deprecated and will be removed in a future release of LightGBM. Pass 'early_stopping()' callback via 'callbacks' argument instead.


[LightGBM] [Warning] Auto-choosing row-wise multi-threading, the overhead of testing was 0.007354 seconds.
You can set `force_row_wise=true` to remove the overhead.
And if memory is not enough, you can set `force_col_wise=true`.
[LightGBM] [Info] Total Bins 867
[LightGBM] [Info] Number of data points in the train set: 185932, number of used features: 11
[LightGBM] [Info] Start training from score 4711.292838
[1]	valid_0's rmse: 4273.7
Training until validation scores don't improve for 500 rounds
[2]	valid_0's rmse: 3988.64
[3]	valid_0's rmse: 3742.02
[4]	valid_0's rmse: 3515
[5]	valid_0's rmse: 3315.85
[6]	valid_0's rmse: 3137.51
[7]	valid_0's rmse: 2977.68
[8]	valid_0's rmse: 2843.18
[9]	valid_0's rmse: 2720.07
[10]	valid_0's rmse: 2613.68
[11]	valid_0's rmse: 2517.57
[12]	valid_0's rmse: 2432.85
[13]	valid_0's rmse: 2361.99
[14]	valid_0's rmse: 2295.44
[15]	valid_0's rmse: 2240.26

C:\Users\XIX\anaconda3\lib\site-packages\lightgbm\basic.py:1780: UserWarning:

Overriding the parameters from Reference Dataset.

C:\Users\XIX\anaconda3\lib\site-packages\lightgbm\basic.py:1513: UserWarning:

categorical_column in param dict is overridden.


[16]	valid_0's rmse: 2189.31
[17]	valid_0's rmse: 2146.01
[18]	valid_0's rmse: 2106.08
[19]	valid_0's rmse: 2072.95
[20]	valid_0's rmse: 2043.83
[21]	valid_0's rmse: 2019.94
[22]	valid_0's rmse: 1996.46
[23]	valid_0's rmse: 1974.5
[24]	valid_0's rmse: 1956.03
[25]	valid_0's rmse: 1939.81
[26]	valid_0's rmse: 1923.95
[27]	valid_0's rmse: 1910.3
[28]	valid_0's rmse: 1898.95
[29]	valid_0's rmse: 1888.57
[30]	valid_0's rmse: 1879.09
[31]	valid_0's rmse: 1871.05
[32]	valid_0's rmse: 1861.02
[33]	valid_0's rmse: 1852.65
[34]	valid_0's rmse: 1846.34
[35]	valid_0's rmse: 1840.52
[36]	valid_0's rmse: 1834.94
[37]	valid_0's rmse: 1830.04
[38]	valid_0's rmse: 1825.7
[39]	valid_0's rmse: 1821.39
[40]	valid_0's rmse: 1817.51
[41]	valid_0's rmse: 1814.52
[42]	valid_0's rmse: 1809.47
[43]	valid_0's rmse: 1805.37
[44]	valid_0's rmse: 1802.84
[45]	valid_0's rmse: 1800.37
[46]	valid_0's rmse: 1797.9
[47]	valid_0's rmse: 1796.13
[48]	valid_0's rmse: 1794.19
[49]	valid_0's rmse: 1791.32
[50]	valid_0's rmse: 1790.11
[51]	valid_0's rmse: 1788.18
[52]	valid_0's rmse: 1786.29
[53]	valid_0's rmse: 1784.35
[54]	valid_0's rmse: 1783.24
[55]	valid_0's rmse: 1781.54
[56]	valid_0's rmse: 1779.8
[57]	valid_0's rmse: 1778.22
[58]	valid_0's rmse: 1776.91
[59]	valid_0's rmse: 1776.09
[60]	valid_0's rmse: 1774.35
[61]	valid_0's rmse: 1773.26
[62]	valid_0's rmse: 1772.58
[63]	valid_0's rmse: 1770.98
[64]	valid_0's rmse: 1769.45
[65]	valid_0's rmse: 1768.12
[66]	valid_0's rmse: 1767.39
[67]	valid_0's rmse: 1766.34
[68]	valid_0's rmse: 1764.53
[69]	valid_0's rmse: 1763.05
[70]	valid_0's rmse: 1761.34
[71]	valid_0's rmse: 1760.57
[72]	valid_0's rmse: 1759.85
[73]	valid_0's rmse: 1759.41
[74]	valid_0's rmse: 1758.21
[75]	valid_0's rmse: 1757.37
[76]	valid_0's rmse: 1755.74
[77]	valid_0's rmse: 1754.52
[78]	valid_0's rmse: 1754.27
[79]	valid_0's rmse: 1752.54
[80]	valid_0's rmse: 1751.45
[81]	valid_0's rmse: 1750.31
[82]	valid_0's rmse: 1749.43
[83]	valid_0's rmse: 1748.28
[84]	valid_0's rmse: 1747.59
[85]	valid_0's rmse: 1747.18
[86]	valid_0's rmse: 1746.62
[87]	valid_0's rmse: 1745.85
[88]	valid_0's rmse: 1744.68
[89]	valid_0's rmse: 1744.01
[90]	valid_0's rmse: 1743.42
[91]	valid_0's rmse: 1742.65
[92]	valid_0's rmse: 1741.63
[93]	valid_0's rmse: 1740.93
[94]	valid_0's rmse: 1740.67
[95]	valid_0's rmse: 1740.04
[96]	valid_0's rmse: 1739.17
[97]	valid_0's rmse: 1738.46
[98]	valid_0's rmse: 1738.01
[99]	valid_0's rmse: 1737.64
[100]	valid_0's rmse: 1737.52
[101]	valid_0's rmse: 1737.08
[102]	valid_0's rmse: 1735.65
[103]	valid_0's rmse: 1735.26
[104]	valid_0's rmse: 1735
[105]	valid_0's rmse: 1734.67
[106]	valid_0's rmse: 1734.37
[107]	valid_0's rmse: 1734.09
[108]	valid_0's rmse: 1733.71
[109]	valid_0's rmse: 1733.34
[110]	valid_0's rmse: 1732.85
[111]	valid_0's rmse: 1732.61
[112]	valid_0's rmse: 1732.26
[113]	valid_0's rmse: 1731.98
[114]	valid_0's rmse: 1731.68
[115]	valid_0's rmse: 1731.63
[116]	valid_0's rmse: 1730.84
[117]	valid_0's rmse: 1729.85
[118]	valid_0's rmse: 1729.69
[119]	valid_0's rmse: 1729.44
[120]	valid_0's rmse: 1729.37
[121]	valid_0's rmse: 1728.47
[122]	valid_0's rmse: 1728.18
[123]	valid_0's rmse: 1727.86
[124]	valid_0's rmse: 1727.53
[125]	valid_0's rmse: 1727.48
[126]	valid_0's rmse: 1727.01
[127]	valid_0's rmse: 1726.88
[128]	valid_0's rmse: 1726.94
[129]	valid_0's rmse: 1726.7
[130]	valid_0's rmse: 1726.3
[131]	valid_0's rmse: 1725.2
[132]	valid_0's rmse: 1724.95
[133]	valid_0's rmse: 1724.67
[134]	valid_0's rmse: 1724.37
[135]	valid_0's rmse: 1723.61
[136]	valid_0's rmse: 1722.76
[137]	valid_0's rmse: 1722.62
[138]	valid_0's rmse: 1722.31
[139]	valid_0's rmse: 1721.34
[140]	valid_0's rmse: 1721.03
[141]	valid_0's rmse: 1720.39
[142]	valid_0's rmse: 1720.35
[143]	valid_0's rmse: 1720.17
[144]	valid_0's rmse: 1720.14
[145]	valid_0's rmse: 1720
[146]	valid_0's rmse: 1719.84
[147]	valid_0's rmse: 1719.6
[148]	valid_0's rmse: 1719.49
[149]	valid_0's rmse: 1718.85
[150]	valid_0's rmse: 1718.76
[151]	valid_0's rmse: 1718.61
[152]	valid_0's rmse: 1718.04
[153]	valid_0's rmse: 1717.37
[154]	valid_0's rmse: 1717.16
[155]	valid_0's rmse: 1717.12
[156]	valid_0's rmse: 1717.03
[157]	valid_0's rmse: 1716.99
[158]	valid_0's rmse: 1716.04
[159]	valid_0's rmse: 1715.99
[160]	valid_0's rmse: 1715.89
[161]	valid_0's rmse: 1716.05
[162]	valid_0's rmse: 1715.74
[163]	valid_0's rmse: 1715.36
[164]	valid_0's rmse: 1715.04
[165]	valid_0's rmse: 1714.78
[166]	valid_0's rmse: 1714.63
[167]	valid_0's rmse: 1714.45
[168]	valid_0's rmse: 1714.39
[169]	valid_0's rmse: 1714.5
[170]	valid_0's rmse: 1714.46
[171]	valid_0's rmse: 1713.9
[172]	valid_0's rmse: 1713.76
[173]	valid_0's rmse: 1713.6
[174]	valid_0's rmse: 1713.32
[175]	valid_0's rmse: 1712.91
[176]	valid_0's rmse: 1712.7
[177]	valid_0's rmse: 1712.61
[178]	valid_0's rmse: 1712.38
[179]	valid_0's rmse: 1712.33
[180]	valid_0's rmse: 1712.3
[181]	valid_0's rmse: 1711.97
[182]	valid_0's rmse: 1711.3
[183]	valid_0's rmse: 1711.31
[184]	valid_0's rmse: 1711.09
[185]	valid_0's rmse: 1710.85
[186]	valid_0's rmse: 1710.8
[187]	valid_0's rmse: 1710.58
[188]	valid_0's rmse: 1710.47
[189]	valid_0's rmse: 1710.18
[190]	valid_0's rmse: 1709.69
[191]	valid_0's rmse: 1709.06
[192]	valid_0's rmse: 1708.89
[193]	valid_0's rmse: 1708.39
[194]	valid_0's rmse: 1708.36
[195]	valid_0's rmse: 1708.37
[196]	valid_0's rmse: 1708.1
[197]	valid_0's rmse: 1708.03
[198]	valid_0's rmse: 1707.83
[199]	valid_0's rmse: 1707.51
[200]	valid_0's rmse: 1707.34
[201]	valid_0's rmse: 1707.24
[202]	valid_0's rmse: 1706.81
[203]	valid_0's rmse: 1706.81
[204]	valid_0's rmse: 1706.63
[205]	valid_0's rmse: 1706.5
[206]	valid_0's rmse: 1705.9
[207]	valid_0's rmse: 1705.83
[208]	valid_0's rmse: 1705.73
[209]	valid_0's rmse: 1705.76
[210]	valid_0's rmse: 1705.41
[211]	valid_0's rmse: 1704.88
[212]	valid_0's rmse: 1704.75
[213]	valid_0's rmse: 1704.52
[214]	valid_0's rmse: 1704.15
[215]	valid_0's rmse: 1704.24
[216]	valid_0's rmse: 1703.98
[217]	valid_0's rmse: 1703.83
[218]	valid_0's rmse: 1703.76
[219]	valid_0's rmse: 1703.53
[220]	valid_0's rmse: 1703.47
[221]	valid_0's rmse: 1703.26
[222]	valid_0's rmse: 1702.94
[223]	valid_0's rmse: 1702.73
[224]	valid_0's rmse: 1702.64
[225]	valid_0's rmse: 1702.59
[226]	valid_0's rmse: 1702.48
[227]	valid_0's rmse: 1702.14
[228]	valid_0's rmse: 1702.17
[229]	valid_0's rmse: 1702.06
[230]	valid_0's rmse: 1701.88
[231]	valid_0's rmse: 1701.85
[232]	valid_0's rmse: 1701.83
[233]	valid_0's rmse: 1701.63
[234]	valid_0's rmse: 1701.6
[235]	valid_0's rmse: 1701.37
[236]	valid_0's rmse: 1701.32
[237]	valid_0's rmse: 1701.31
[238]	valid_0's rmse: 1701.37
[239]	valid_0's rmse: 1701.26
[240]	valid_0's rmse: 1701.07
[241]	valid_0's rmse: 1701.21
[242]	valid_0's rmse: 1700.95
[243]	valid_0's rmse: 1701.03
[244]	valid_0's rmse: 1700.8
[245]	valid_0's rmse: 1700.77
[246]	valid_0's rmse: 1700.67
[247]	valid_0's rmse: 1700.55
[248]	valid_0's rmse: 1700.39
[249]	valid_0's rmse: 1700.38
[250]	valid_0's rmse: 1700.1
[251]	valid_0's rmse: 1700.06
[252]	valid_0's rmse: 1700.08
[253]	valid_0's rmse: 1699.91
[254]	valid_0's rmse: 1699.93
[255]	valid_0's rmse: 1699.9
[256]	valid_0's rmse: 1699.82
[257]	valid_0's rmse: 1699.72
[258]	valid_0's rmse: 1699.6
[259]	valid_0's rmse: 1699.3
[260]	valid_0's rmse: 1699.15
[261]	valid_0's rmse: 1699.03
[262]	valid_0's rmse: 1698.73
[263]	valid_0's rmse: 1698.61
[264]	valid_0's rmse: 1698.38
[265]	valid_0's rmse: 1698.33
[266]	valid_0's rmse: 1698.36
[267]	valid_0's rmse: 1698.39
[268]	valid_0's rmse: 1698
[269]	valid_0's rmse: 1697.99
[270]	valid_0's rmse: 1697.79
[271]	valid_0's rmse: 1697.8
[272]	valid_0's rmse: 1697.69
[273]	valid_0's rmse: 1697.67
[274]	valid_0's rmse: 1697.65
[275]	valid_0's rmse: 1697.55
[276]	valid_0's rmse: 1697.28
[277]	valid_0's rmse: 1697.26
[278]	valid_0's rmse: 1697.29
[279]	valid_0's rmse: 1697.23
[280]	valid_0's rmse: 1697.1
[281]	valid_0's rmse: 1697.1
[282]	valid_0's rmse: 1696.87
[283]	valid_0's rmse: 1696.89
[284]	valid_0's rmse: 1696.68
[285]	valid_0's rmse: 1696.64
[286]	valid_0's rmse: 1696.34
[287]	valid_0's rmse: 1696.28
[288]	valid_0's rmse: 1695.82
[289]	valid_0's rmse: 1695.56
[290]	valid_0's rmse: 1695.53
[291]	valid_0's rmse: 1695.55
[292]	valid_0's rmse: 1695.44
[293]	valid_0's rmse: 1695.38
[294]	valid_0's rmse: 1695.29
[295]	valid_0's rmse: 1695.21
[296]	valid_0's rmse: 1695.14
[297]	valid_0's rmse: 1695.05
[298]	valid_0's rmse: 1694.81
[299]	valid_0's rmse: 1694.78
[300]	valid_0's rmse: 1694.81
[301]	valid_0's rmse: 1694.29
[302]	valid_0's rmse: 1694.17
[303]	valid_0's rmse: 1693.73
[304]	valid_0's rmse: 1693.6
[305]	valid_0's rmse: 1693.52
[306]	valid_0's rmse: 1693.31
[307]	valid_0's rmse: 1693.4
[308]	valid_0's rmse: 1693.35
[309]	valid_0's rmse: 1693.3
[310]	valid_0's rmse: 1693.25
[311]	valid_0's rmse: 1693.2
[312]	valid_0's rmse: 1693.27
[313]	valid_0's rmse: 1693.3
[314]	valid_0's rmse: 1693.07
[315]	valid_0's rmse: 1692.9
[316]	valid_0's rmse: 1692.78
[317]	valid_0's rmse: 1692.69
[318]	valid_0's rmse: 1692.56
[319]	valid_0's rmse: 1692.56
[320]	valid_0's rmse: 1692.41
[321]	valid_0's rmse: 1692.39
[322]	valid_0's rmse: 1692.21
[323]	valid_0's rmse: 1692.02
[324]	valid_0's rmse: 1692
[325]	valid_0's rmse: 1692.05
[326]	valid_0's rmse: 1692.1
[327]	valid_0's rmse: 1692.04
[328]	valid_0's rmse: 1691.95
[329]	valid_0's rmse: 1691.91
[330]	valid_0's rmse: 1691.91
[331]	valid_0's rmse: 1691.54
[332]	valid_0's rmse: 1690.97
[333]	valid_0's rmse: 1690.89
[334]	valid_0's rmse: 1690.63
[335]	valid_0's rmse: 1690.37
[336]	valid_0's rmse: 1690.38
[337]	valid_0's rmse: 1690.37
[338]	valid_0's rmse: 1690.35
[339]	valid_0's rmse: 1690.12
[340]	valid_0's rmse: 1690.19
[341]	valid_0's rmse: 1689.38
[342]	valid_0's rmse: 1689.29
[343]	valid_0's rmse: 1689.1
[344]	valid_0's rmse: 1689.12
[345]	valid_0's rmse: 1688.99
[346]	valid_0's rmse: 1688.82
[347]	valid_0's rmse: 1688.61
[348]	valid_0's rmse: 1688.43
[349]	valid_0's rmse: 1688.41
[350]	valid_0's rmse: 1688.47
[351]	valid_0's rmse: 1688.48
[352]	valid_0's rmse: 1688.04
[353]	valid_0's rmse: 1688.04
[354]	valid_0's rmse: 1687.91
[355]	valid_0's rmse: 1687.92
[356]	valid_0's rmse: 1687.87
[357]	valid_0's rmse: 1687.88
[358]	valid_0's rmse: 1687.85
[359]	valid_0's rmse: 1687.81
[360]	valid_0's rmse: 1687.72
[361]	valid_0's rmse: 1687.68
[362]	valid_0's rmse: 1687.73
[363]	valid_0's rmse: 1687.78
[364]	valid_0's rmse: 1687.7
[365]	valid_0's rmse: 1687.63
[366]	valid_0's rmse: 1687.41
[367]	valid_0's rmse: 1686.86
[368]	valid_0's rmse: 1686.83
[369]	valid_0's rmse: 1686.63
[370]	valid_0's rmse: 1686.68
[371]	valid_0's rmse: 1686.57
[372]	valid_0's rmse: 1686.49
[373]	valid_0's rmse: 1686.41
[374]	valid_0's rmse: 1686.37
[375]	valid_0's rmse: 1686.41
[376]	valid_0's rmse: 1686.3
[377]	valid_0's rmse: 1686.26
[378]	valid_0's rmse: 1686.33
[379]	valid_0's rmse: 1686.3
[380]	valid_0's rmse: 1685.99
[381]	valid_0's rmse: 1685.93
[382]	valid_0's rmse: 1685.69
[383]	valid_0's rmse: 1685.45
[384]	valid_0's rmse: 1685.12
[385]	valid_0's rmse: 1684.81
[386]	valid_0's rmse: 1684.81
[387]	valid_0's rmse: 1684.66
[388]	valid_0's rmse: 1684.55
[389]	valid_0's rmse: 1684.48
[390]	valid_0's rmse: 1684.5
[391]	valid_0's rmse: 1684.25
[392]	valid_0's rmse: 1684.08
[393]	valid_0's rmse: 1684.12
[394]	valid_0's rmse: 1683.93
[395]	valid_0's rmse: 1683.94
[396]	valid_0's rmse: 1683.79
[397]	valid_0's rmse: 1683.65
[398]	valid_0's rmse: 1683.65
[399]	valid_0's rmse: 1683.64
[400]	valid_0's rmse: 1683.58
[401]	valid_0's rmse: 1683.43
[402]	valid_0's rmse: 1683.47
[403]	valid_0's rmse: 1683.34
[404]	valid_0's rmse: 1683.17
[405]	valid_0's rmse: 1683.12
[406]	valid_0's rmse: 1682.72
[407]	valid_0's rmse: 1682.6
[408]	valid_0's rmse: 1682.46
[409]	valid_0's rmse: 1682.28
[410]	valid_0's rmse: 1682.21
[411]	valid_0's rmse: 1681.92
[412]	valid_0's rmse: 1681.9
[413]	valid_0's rmse: 1681.6
[414]	valid_0's rmse: 1681.52
[415]	valid_0's rmse: 1681.61
[416]	valid_0's rmse: 1681.6
[417]	valid_0's rmse: 1681.44
[418]	valid_0's rmse: 1681.5
[419]	valid_0's rmse: 1681.3
[420]	valid_0's rmse: 1681.31
[421]	valid_0's rmse: 1681.36
[422]	valid_0's rmse: 1681.41
[423]	valid_0's rmse: 1681.45
[424]	valid_0's rmse: 1681.32
[425]	valid_0's rmse: 1681.17
[426]	valid_0's rmse: 1681.17
[427]	valid_0's rmse: 1681.24
[428]	valid_0's rmse: 1681.19
[429]	valid_0's rmse: 1680.97
[430]	valid_0's rmse: 1680.97
[431]	valid_0's rmse: 1680.9
[432]	valid_0's rmse: 1680.8
[433]	valid_0's rmse: 1680.74
[434]	valid_0's rmse: 1680.65
[435]	valid_0's rmse: 1680.54
[436]	valid_0's rmse: 1680.58
[437]	valid_0's rmse: 1680.57
[438]	valid_0's rmse: 1680.44
[439]	valid_0's rmse: 1680.42
[440]	valid_0's rmse: 1680.52
[441]	valid_0's rmse: 1680.47
[442]	valid_0's rmse: 1680.38
[443]	valid_0's rmse: 1680.43
[444]	valid_0's rmse: 1680.3
[445]	valid_0's rmse: 1680.24
[446]	valid_0's rmse: 1680.19
[447]	valid_0's rmse: 1680.21
[448]	valid_0's rmse: 1680.25
[449]	valid_0's rmse: 1680.16
[450]	valid_0's rmse: 1680.15
[451]	valid_0's rmse: 1679.97
[452]	valid_0's rmse: 1679.6
[453]	valid_0's rmse: 1679.62
[454]	valid_0's rmse: 1679.5
[455]	valid_0's rmse: 1679.3
[456]	valid_0's rmse: 1679.31
[457]	valid_0's rmse: 1679.35
[458]	valid_0's rmse: 1679.29
[459]	valid_0's rmse: 1678.94
[460]	valid_0's rmse: 1678.81
[461]	valid_0's rmse: 1678.73
[462]	valid_0's rmse: 1678.63
[463]	valid_0's rmse: 1678.54
[464]	valid_0's rmse: 1678.42
[465]	valid_0's rmse: 1678.39
[466]	valid_0's rmse: 1678.12
[467]	valid_0's rmse: 1678.21
[468]	valid_0's rmse: 1678.06
[469]	valid_0's rmse: 1677.97
[470]	valid_0's rmse: 1678.07
[471]	valid_0's rmse: 1677.94
[472]	valid_0's rmse: 1677.82
[473]	valid_0's rmse: 1677.64
[474]	valid_0's rmse: 1677.67
[475]	valid_0's rmse: 1677.27
[476]	valid_0's rmse: 1676.89
[477]	valid_0's rmse: 1676.78
[478]	valid_0's rmse: 1676.68
[479]	valid_0's rmse: 1676.51
[480]	valid_0's rmse: 1676.36
[481]	valid_0's rmse: 1676.39
[482]	valid_0's rmse: 1676.19
[483]	valid_0's rmse: 1676.13
[484]	valid_0's rmse: 1676.06
[485]	valid_0's rmse: 1675.97
[486]	valid_0's rmse: 1675.77
[487]	valid_0's rmse: 1675.7
[488]	valid_0's rmse: 1675.7
[489]	valid_0's rmse: 1675.66
[490]	valid_0's rmse: 1675.7
[491]	valid_0's rmse: 1675.68
[492]	valid_0's rmse: 1675.69
[493]	valid_0's rmse: 1675.72
[494]	valid_0's rmse: 1675.69
[495]	valid_0's rmse: 1675.37
[496]	valid_0's rmse: 1675.11
[497]	valid_0's rmse: 1674.99
[498]	valid_0's rmse: 1674.95
[499]	valid_0's rmse: 1674.92
[500]	valid_0's rmse: 1674.74
[501]	valid_0's rmse: 1674.59
[502]	valid_0's rmse: 1674.61
[503]	valid_0's rmse: 1674.58
[504]	valid_0's rmse: 1674.4
[505]	valid_0's rmse: 1674.28
[506]	valid_0's rmse: 1674.28
[507]	valid_0's rmse: 1674.09
[508]	valid_0's rmse: 1673.98
[509]	valid_0's rmse: 1674.06
[510]	valid_0's rmse: 1673.99
[511]	valid_0's rmse: 1673.79
[512]	valid_0's rmse: 1673.83
[513]	valid_0's rmse: 1673.76
[514]	valid_0's rmse: 1673.75
[515]	valid_0's rmse: 1673.92
[516]	valid_0's rmse: 1673.83
[517]	valid_0's rmse: 1673.77
[518]	valid_0's rmse: 1673.74
[519]	valid_0's rmse: 1673.71
[520]	valid_0's rmse: 1673.61
[521]	valid_0's rmse: 1673.57
[522]	valid_0's rmse: 1673.54
[523]	valid_0's rmse: 1673.56
[524]	valid_0's rmse: 1673.54
[525]	valid_0's rmse: 1673.47
[526]	valid_0's rmse: 1673.43
[527]	valid_0's rmse: 1673.44
[528]	valid_0's rmse: 1673.39
[529]	valid_0's rmse: 1673.25
[530]	valid_0's rmse: 1673.16
[531]	valid_0's rmse: 1673.11
[532]	valid_0's rmse: 1673.05
[533]	valid_0's rmse: 1673.07
[534]	valid_0's rmse: 1673
[535]	valid_0's rmse: 1672.79
[536]	valid_0's rmse: 1672.52
[537]	valid_0's rmse: 1672.39
[538]	valid_0's rmse: 1672.43
[539]	valid_0's rmse: 1672.37
[540]	valid_0's rmse: 1672.21
[541]	valid_0's rmse: 1672.18
[542]	valid_0's rmse: 1672.19
[543]	valid_0's rmse: 1672.21
[544]	valid_0's rmse: 1672.26
[545]	valid_0's rmse: 1672.17
[546]	valid_0's rmse: 1672.18
[547]	valid_0's rmse: 1672.04
[548]	valid_0's rmse: 1672.03
[549]	valid_0's rmse: 1671.98
[550]	valid_0's rmse: 1672
[551]	valid_0's rmse: 1671.91
[552]	valid_0's rmse: 1671.78
[553]	valid_0's rmse: 1671.7
[554]	valid_0's rmse: 1671.69
[555]	valid_0's rmse: 1671.67
[556]	valid_0's rmse: 1671.65
[557]	valid_0's rmse: 1671.68
[558]	valid_0's rmse: 1671.65
[559]	valid_0's rmse: 1671.65
[560]	valid_0's rmse: 1671.55
[561]	valid_0's rmse: 1671.42
[562]	valid_0's rmse: 1671.46
[563]	valid_0's rmse: 1671.35
[564]	valid_0's rmse: 1671.34
[565]	valid_0's rmse: 1671.22
[566]	valid_0's rmse: 1671.17
[567]	valid_0's rmse: 1671.13
[568]	valid_0's rmse: 1671.1
[569]	valid_0's rmse: 1671.1
[570]	valid_0's rmse: 1670.96
[571]	valid_0's rmse: 1670.95
[572]	valid_0's rmse: 1670.91
[573]	valid_0's rmse: 1671.03
[574]	valid_0's rmse: 1670.94
[575]	valid_0's rmse: 1670.91
[576]	valid_0's rmse: 1670.85
[577]	valid_0's rmse: 1670.85
[578]	valid_0's rmse: 1670.86
[579]	valid_0's rmse: 1670.87
[580]	valid_0's rmse: 1670.72
[581]	valid_0's rmse: 1670.7
[582]	valid_0's rmse: 1670.53
[583]	valid_0's rmse: 1670.54
[584]	valid_0's rmse: 1670.46
[585]	valid_0's rmse: 1670.5
[586]	valid_0's rmse: 1670.54
[587]	valid_0's rmse: 1670.39
[588]	valid_0's rmse: 1670.28
[589]	valid_0's rmse: 1670.18
[590]	valid_0's rmse: 1670.14
[591]	valid_0's rmse: 1670.16
[592]	valid_0's rmse: 1670.1
[593]	valid_0's rmse: 1669.99
[594]	valid_0's rmse: 1669.9
[595]	valid_0's rmse: 1669.84
[596]	valid_0's rmse: 1669.8
[597]	valid_0's rmse: 1669.78
[598]	valid_0's rmse: 1669.63
[599]	valid_0's rmse: 1669.52
[600]	valid_0's rmse: 1669.37
[601]	valid_0's rmse: 1669.36
[602]	valid_0's rmse: 1669.29
[603]	valid_0's rmse: 1669.27
[604]	valid_0's rmse: 1669.17
[605]	valid_0's rmse: 1669.12
[606]	valid_0's rmse: 1669.23
[607]	valid_0's rmse: 1669.21
[608]	valid_0's rmse: 1669.13
[609]	valid_0's rmse: 1669.05
[610]	valid_0's rmse: 1668.98
[611]	valid_0's rmse: 1668.95
[612]	valid_0's rmse: 1668.87
[613]	valid_0's rmse: 1668.86
[614]	valid_0's rmse: 1668.87
[615]	valid_0's rmse: 1668.96
[616]	valid_0's rmse: 1668.94
[617]	valid_0's rmse: 1668.92
[618]	valid_0's rmse: 1668.98
[619]	valid_0's rmse: 1668.9
[620]	valid_0's rmse: 1668.9
[621]	valid_0's rmse: 1668.81
[622]	valid_0's rmse: 1668.7
[623]	valid_0's rmse: 1668.43
[624]	valid_0's rmse: 1668.45
[625]	valid_0's rmse: 1668.4
[626]	valid_0's rmse: 1668.01
[627]	valid_0's rmse: 1668
[628]	valid_0's rmse: 1667.7
[629]	valid_0's rmse: 1667.71
[630]	valid_0's rmse: 1667.51
[631]	valid_0's rmse: 1667.4
[632]	valid_0's rmse: 1667.51
[633]	valid_0's rmse: 1667.43
[634]	valid_0's rmse: 1667.21
[635]	valid_0's rmse: 1667.23
[636]	valid_0's rmse: 1667.21
[637]	valid_0's rmse: 1667.2
[638]	valid_0's rmse: 1667.15
[639]	valid_0's rmse: 1667.13
[640]	valid_0's rmse: 1667.09
[641]	valid_0's rmse: 1667.17
[642]	valid_0's rmse: 1667.17
[643]	valid_0's rmse: 1667.13
[644]	valid_0's rmse: 1667.02
[645]	valid_0's rmse: 1666.91
[646]	valid_0's rmse: 1666.84
[647]	valid_0's rmse: 1666.84
[648]	valid_0's rmse: 1666.84
[649]	valid_0's rmse: 1666.73
[650]	valid_0's rmse: 1666.73
[651]	valid_0's rmse: 1666.46
[652]	valid_0's rmse: 1666.46
[653]	valid_0's rmse: 1666.5
[654]	valid_0's rmse: 1666.47
[655]	valid_0's rmse: 1666.41
[656]	valid_0's rmse: 1666.45
[657]	valid_0's rmse: 1666.44
[658]	valid_0's rmse: 1666.37
[659]	valid_0's rmse: 1666.37
[660]	valid_0's rmse: 1666.36
[661]	valid_0's rmse: 1666.28
[662]	valid_0's rmse: 1666.3
[663]	valid_0's rmse: 1666.24
[664]	valid_0's rmse: 1666.23
[665]	valid_0's rmse: 1666.18
[666]	valid_0's rmse: 1666.18
[667]	valid_0's rmse: 1665.99
[668]	valid_0's rmse: 1665.95
[669]	valid_0's rmse: 1665.91
[670]	valid_0's rmse: 1665.88
[671]	valid_0's rmse: 1665.83
[672]	valid_0's rmse: 1665.84
[673]	valid_0's rmse: 1665.65
[674]	valid_0's rmse: 1665.63
[675]	valid_0's rmse: 1665.58
[676]	valid_0's rmse: 1665.5
[677]	valid_0's rmse: 1665.53
[678]	valid_0's rmse: 1665.32
[679]	valid_0's rmse: 1665.15
[680]	valid_0's rmse: 1665.09
[681]	valid_0's rmse: 1665.04
[682]	valid_0's rmse: 1664.97
[683]	valid_0's rmse: 1664.82
[684]	valid_0's rmse: 1664.66
[685]	valid_0's rmse: 1664.2
[686]	valid_0's rmse: 1664.17
[687]	valid_0's rmse: 1664.01
[688]	valid_0's rmse: 1663.97
[689]	valid_0's rmse: 1663.83
[690]	valid_0's rmse: 1663.74
[691]	valid_0's rmse: 1663.58
[692]	valid_0's rmse: 1663.54
[693]	valid_0's rmse: 1663.44
[694]	valid_0's rmse: 1663.41
[695]	valid_0's rmse: 1663.41
[696]	valid_0's rmse: 1663.29
[697]	valid_0's rmse: 1663.12
[698]	valid_0's rmse: 1663.11
[699]	valid_0's rmse: 1663.05
[700]	valid_0's rmse: 1663
[701]	valid_0's rmse: 1662.92
[702]	valid_0's rmse: 1662.93
[703]	valid_0's rmse: 1662.99
[704]	valid_0's rmse: 1662.93
[705]	valid_0's rmse: 1662.84
[706]	valid_0's rmse: 1662.83
[707]	valid_0's rmse: 1662.83
[708]	valid_0's rmse: 1662.8
[709]	valid_0's rmse: 1662.81
[710]	valid_0's rmse: 1662.8
[711]	valid_0's rmse: 1662.74
[712]	valid_0's rmse: 1662.9
[713]	valid_0's rmse: 1662.92
[714]	valid_0's rmse: 1662.79
[715]	valid_0's rmse: 1662.6
[716]	valid_0's rmse: 1662.56
[717]	valid_0's rmse: 1662.56
[718]	valid_0's rmse: 1662.53
[719]	valid_0's rmse: 1662.52
[720]	valid_0's rmse: 1662.44
[721]	valid_0's rmse: 1662.4
[722]	valid_0's rmse: 1662.32
[723]	valid_0's rmse: 1662.38
[724]	valid_0's rmse: 1662.42
[725]	valid_0's rmse: 1662.37
[726]	valid_0's rmse: 1662.31
[727]	valid_0's rmse: 1662.3
[728]	valid_0's rmse: 1662.31
[729]	valid_0's rmse: 1662.29
[730]	valid_0's rmse: 1662.27
[731]	valid_0's rmse: 1662.26
[732]	valid_0's rmse: 1662.2
[733]	valid_0's rmse: 1662.17
[734]	valid_0's rmse: 1662.19
[735]	valid_0's rmse: 1662.19
[736]	valid_0's rmse: 1662.22
[737]	valid_0's rmse: 1662.25
[738]	valid_0's rmse: 1662.07
[739]	valid_0's rmse: 1661.95
[740]	valid_0's rmse: 1661.93
[741]	valid_0's rmse: 1661.85
[742]	valid_0's rmse: 1661.83
[743]	valid_0's rmse: 1661.68
[744]	valid_0's rmse: 1661.64
[745]	valid_0's rmse: 1661.34
[746]	valid_0's rmse: 1661.4
[747]	valid_0's rmse: 1661.43
[748]	valid_0's rmse: 1661.46
[749]	valid_0's rmse: 1661.37
[750]	valid_0's rmse: 1661.16
[751]	valid_0's rmse: 1661.14
[752]	valid_0's rmse: 1661.1
[753]	valid_0's rmse: 1661.23
[754]	valid_0's rmse: 1661.13
[755]	valid_0's rmse: 1661.09
[756]	valid_0's rmse: 1660.92
[757]	valid_0's rmse: 1660.95
[758]	valid_0's rmse: 1660.95
[759]	valid_0's rmse: 1660.93
[760]	valid_0's rmse: 1660.91
[761]	valid_0's rmse: 1660.86
[762]	valid_0's rmse: 1660.75
[763]	valid_0's rmse: 1660.63
[764]	valid_0's rmse: 1660.62
[765]	valid_0's rmse: 1660.68
[766]	valid_0's rmse: 1660.59
[767]	valid_0's rmse: 1660.44
[768]	valid_0's rmse: 1660.42
[769]	valid_0's rmse: 1660.39
[770]	valid_0's rmse: 1660.19
[771]	valid_0's rmse: 1660.15
[772]	valid_0's rmse: 1660.11
[773]	valid_0's rmse: 1660.08
[774]	valid_0's rmse: 1660.06
[775]	valid_0's rmse: 1660
[776]	valid_0's rmse: 1659.93
[777]	valid_0's rmse: 1659.87
[778]	valid_0's rmse: 1659.82
[779]	valid_0's rmse: 1659.8
[780]	valid_0's rmse: 1659.74
[781]	valid_0's rmse: 1659.78
[782]	valid_0's rmse: 1659.85
[783]	valid_0's rmse: 1659.84
[784]	valid_0's rmse: 1659.81
[785]	valid_0's rmse: 1659.51
[786]	valid_0's rmse: 1659.54
[787]	valid_0's rmse: 1659.49
[788]	valid_0's rmse: 1659.4
[789]	valid_0's rmse: 1659.25
[790]	valid_0's rmse: 1659.24
[791]	valid_0's rmse: 1659.23
[792]	valid_0's rmse: 1659.2
[793]	valid_0's rmse: 1659.03
[794]	valid_0's rmse: 1658.93
[795]	valid_0's rmse: 1658.76
[796]	valid_0's rmse: 1658.74
[797]	valid_0's rmse: 1658.68
[798]	valid_0's rmse: 1658.65
[799]	valid_0's rmse: 1658.67
[800]	valid_0's rmse: 1658.69
[801]	valid_0's rmse: 1658.68
[802]	valid_0's rmse: 1658.7
[803]	valid_0's rmse: 1658.69
[804]	valid_0's rmse: 1658.74
[805]	valid_0's rmse: 1658.68
[806]	valid_0's rmse: 1658.65
[807]	valid_0's rmse: 1658.65
[808]	valid_0's rmse: 1658.62
[809]	valid_0's rmse: 1658.48
[810]	valid_0's rmse: 1658.46
[811]	valid_0's rmse: 1658.38
[812]	valid_0's rmse: 1658.32
[813]	valid_0's rmse: 1658.31
[814]	valid_0's rmse: 1658.25
[815]	valid_0's rmse: 1658.11
[816]	valid_0's rmse: 1658.12
[817]	valid_0's rmse: 1658.11
[818]	valid_0's rmse: 1658.15
[819]	valid_0's rmse: 1658.03
[820]	valid_0's rmse: 1657.91
[821]	valid_0's rmse: 1657.97
[822]	valid_0's rmse: 1657.88
[823]	valid_0's rmse: 1657.89
[824]	valid_0's rmse: 1657.88
[825]	valid_0's rmse: 1657.9
[826]	valid_0's rmse: 1657.91
[827]	valid_0's rmse: 1657.84
[828]	valid_0's rmse: 1657.71
[829]	valid_0's rmse: 1657.58
[830]	valid_0's rmse: 1657.56
[831]	valid_0's rmse: 1657.58
[832]	valid_0's rmse: 1657.47
[833]	valid_0's rmse: 1657.47
[834]	valid_0's rmse: 1657.57
[835]	valid_0's rmse: 1657.47
[836]	valid_0's rmse: 1657.47
[837]	valid_0's rmse: 1657.45
[838]	valid_0's rmse: 1657.38
[839]	valid_0's rmse: 1657.25
[840]	valid_0's rmse: 1657.26
[841]	valid_0's rmse: 1657.38
[842]	valid_0's rmse: 1657.37
[843]	valid_0's rmse: 1657.38
[844]	valid_0's rmse: 1657.22
[845]	valid_0's rmse: 1657.19
[846]	valid_0's rmse: 1657.16
[847]	valid_0's rmse: 1657.06
[848]	valid_0's rmse: 1657.08
[849]	valid_0's rmse: 1657.1
[850]	valid_0's rmse: 1657.06
[851]	valid_0's rmse: 1657.21
[852]	valid_0's rmse: 1657.15
[853]	valid_0's rmse: 1657.11
[854]	valid_0's rmse: 1657.08
[855]	valid_0's rmse: 1657.14
[856]	valid_0's rmse: 1657.11
[857]	valid_0's rmse: 1656.86
[858]	valid_0's rmse: 1656.72
[859]	valid_0's rmse: 1656.66
[860]	valid_0's rmse: 1656.56
[861]	valid_0's rmse: 1656.52
[862]	valid_0's rmse: 1656.5
[863]	valid_0's rmse: 1656.49
[864]	valid_0's rmse: 1656.5
[865]	valid_0's rmse: 1656.45
[866]	valid_0's rmse: 1656.43
[867]	valid_0's rmse: 1656.39
[868]	valid_0's rmse: 1656.29
[869]	valid_0's rmse: 1656.29
[870]	valid_0's rmse: 1656.24
[871]	valid_0's rmse: 1656.24
[872]	valid_0's rmse: 1656.27
[873]	valid_0's rmse: 1656.25
[874]	valid_0's rmse: 1656.25
[875]	valid_0's rmse: 1656.19
[876]	valid_0's rmse: 1656.06
[877]	valid_0's rmse: 1656.05
[878]	valid_0's rmse: 1656.13
[879]	valid_0's rmse: 1656.12
[880]	valid_0's rmse: 1656.05
[881]	valid_0's rmse: 1656.08
[882]	valid_0's rmse: 1656.06
[883]	valid_0's rmse: 1656.1
[884]	valid_0's rmse: 1656.03
[885]	valid_0's rmse: 1655.95
[886]	valid_0's rmse: 1655.9
[887]	valid_0's rmse: 1655.92
[888]	valid_0's rmse: 1655.93
[889]	valid_0's rmse: 1655.83
[890]	valid_0's rmse: 1655.82
[891]	valid_0's rmse: 1655.81
[892]	valid_0's rmse: 1655.75
[893]	valid_0's rmse: 1655.73
[894]	valid_0's rmse: 1655.78
[895]	valid_0's rmse: 1655.66
[896]	valid_0's rmse: 1655.5
[897]	valid_0's rmse: 1655.56
[898]	valid_0's rmse: 1655.67
[899]	valid_0's rmse: 1655.58
[900]	valid_0's rmse: 1655.63
[901]	valid_0's rmse: 1655.62
[902]	valid_0's rmse: 1655.51
[903]	valid_0's rmse: 1655.44
[904]	valid_0's rmse: 1655.39
[905]	valid_0's rmse: 1655.35
[906]	valid_0's rmse: 1655.32
[907]	valid_0's rmse: 1655.31
[908]	valid_0's rmse: 1655.33
[909]	valid_0's rmse: 1655.27
[910]	valid_0's rmse: 1655.27
[911]	valid_0's rmse: 1655.24
[912]	valid_0's rmse: 1655.25
[913]	valid_0's rmse: 1655.2
[914]	valid_0's rmse: 1655.13
[915]	valid_0's rmse: 1655.19
[916]	valid_0's rmse: 1655.21
[917]	valid_0's rmse: 1655.22
[918]	valid_0's rmse: 1655.23
[919]	valid_0's rmse: 1655.12
[920]	valid_0's rmse: 1655.16
[921]	valid_0's rmse: 1655.1
[922]	valid_0's rmse: 1655.11
[923]	valid_0's rmse: 1655.1
[924]	valid_0's rmse: 1655.11
[925]	valid_0's rmse: 1655.07
[926]	valid_0's rmse: 1654.94
[927]	valid_0's rmse: 1654.94
[928]	valid_0's rmse: 1654.9
[929]	valid_0's rmse: 1654.84
[930]	valid_0's rmse: 1654.82
[931]	valid_0's rmse: 1654.78
[932]	valid_0's rmse: 1654.84
[933]	valid_0's rmse: 1654.84
[934]	valid_0's rmse: 1654.85
[935]	valid_0's rmse: 1654.79
[936]	valid_0's rmse: 1654.8
[937]	valid_0's rmse: 1654.74
[938]	valid_0's rmse: 1654.73
[939]	valid_0's rmse: 1654.64
[940]	valid_0's rmse: 1654.61
[941]	valid_0's rmse: 1654.62
[942]	valid_0's rmse: 1654.61
[943]	valid_0's rmse: 1654.41
[944]	valid_0's rmse: 1654.26
[945]	valid_0's rmse: 1654.22
[946]	valid_0's rmse: 1654.19
[947]	valid_0's rmse: 1654.18
[948]	valid_0's rmse: 1654.14
[949]	valid_0's rmse: 1654.17
[950]	valid_0's rmse: 1654.07
[951]	valid_0's rmse: 1653.99
[952]	valid_0's rmse: 1653.98
[953]	valid_0's rmse: 1654.03
[954]	valid_0's rmse: 1653.96
[955]	valid_0's rmse: 1653.92
[956]	valid_0's rmse: 1653.92
[957]	valid_0's rmse: 1653.9
[958]	valid_0's rmse: 1653.88
[959]	valid_0's rmse: 1653.84
[960]	valid_0's rmse: 1653.8
[961]	valid_0's rmse: 1653.76
[962]	valid_0's rmse: 1653.77
[963]	valid_0's rmse: 1653.73
[964]	valid_0's rmse: 1653.71
[965]	valid_0's rmse: 1653.57
[966]	valid_0's rmse: 1653.44
[967]	valid_0's rmse: 1653.31
[968]	valid_0's rmse: 1653.28
[969]	valid_0's rmse: 1653.26
[970]	valid_0's rmse: 1653.33
[971]	valid_0's rmse: 1653.33
[972]	valid_0's rmse: 1653.26
[973]	valid_0's rmse: 1653.21
[974]	valid_0's rmse: 1653.27
[975]	valid_0's rmse: 1653.27
[976]	valid_0's rmse: 1653.27
[977]	valid_0's rmse: 1653.26
[978]	valid_0's rmse: 1653.29
[979]	valid_0's rmse: 1653.28
[980]	valid_0's rmse: 1653.27
[981]	valid_0's rmse: 1653.26
[982]	valid_0's rmse: 1653.2
[983]	valid_0's rmse: 1653.12
[984]	valid_0's rmse: 1653.11
[985]	valid_0's rmse: 1653.09
[986]	valid_0's rmse: 1653.1
[987]	valid_0's rmse: 1653.09
[988]	valid_0's rmse: 1653.19
[989]	valid_0's rmse: 1653.2
[990]	valid_0's rmse: 1653.18
[991]	valid_0's rmse: 1653.21
[992]	valid_0's rmse: 1653.18
[993]	valid_0's rmse: 1653.18
[994]	valid_0's rmse: 1653.17
[995]	valid_0's rmse: 1653.16
[996]	valid_0's rmse: 1653.17
[997]	valid_0's rmse: 1653.17
[998]	valid_0's rmse: 1653.19
[999]	valid_0's rmse: 1653.14
[1000]	valid_0's rmse: 1653.26
[1001]	valid_0's rmse: 1653.25
[1002]	valid_0's rmse: 1653.25
[1003]	valid_0's rmse: 1653.27
[1004]	valid_0's rmse: 1653.26
[1005]	valid_0's rmse: 1653.25
[1006]	valid_0's rmse: 1653.21
[1007]	valid_0's rmse: 1653.2
[1008]	valid_0's rmse: 1653.21
[1009]	valid_0's rmse: 1653.22
[1010]	valid_0's rmse: 1653.19
[1011]	valid_0's rmse: 1653.11
[1012]	valid_0's rmse: 1653.1
[1013]	valid_0's rmse: 1653.1
[1014]	valid_0's rmse: 1653.04
[1015]	valid_0's rmse: 1653.02
[1016]	valid_0's rmse: 1653.08
[1017]	valid_0's rmse: 1653.08
[1018]	valid_0's rmse: 1653.11
[1019]	valid_0's rmse: 1653.04
[1020]	valid_0's rmse: 1653.03
[1021]	valid_0's rmse: 1652.99
[1022]	valid_0's rmse: 1653.01
[1023]	valid_0's rmse: 1652.89
[1024]	valid_0's rmse: 1652.84
[1025]	valid_0's rmse: 1652.82
[1026]	valid_0's rmse: 1652.76
[1027]	valid_0's rmse: 1652.67
[1028]	valid_0's rmse: 1652.66
[1029]	valid_0's rmse: 1652.57
[1030]	valid_0's rmse: 1652.47
[1031]	valid_0's rmse: 1652.47
[1032]	valid_0's rmse: 1652.45
[1033]	valid_0's rmse: 1652.48
[1034]	valid_0's rmse: 1652.59
[1035]	valid_0's rmse: 1652.53
[1036]	valid_0's rmse: 1652.49
[1037]	valid_0's rmse: 1652.47
[1038]	valid_0's rmse: 1652.44
[1039]	valid_0's rmse: 1652.41
[1040]	valid_0's rmse: 1652.4
[1041]	valid_0's rmse: 1652.41
[1042]	valid_0's rmse: 1652.4
[1043]	valid_0's rmse: 1652.43
[1044]	valid_0's rmse: 1652.33
[1045]	valid_0's rmse: 1652.32
[1046]	valid_0's rmse: 1652.26
[1047]	valid_0's rmse: 1652.26
[1048]	valid_0's rmse: 1652.25
[1049]	valid_0's rmse: 1652.27
[1050]	valid_0's rmse: 1652.13
[1051]	valid_0's rmse: 1652.09
[1052]	valid_0's rmse: 1652.15
[1053]	valid_0's rmse: 1652.16
[1054]	valid_0's rmse: 1652.15
[1055]	valid_0's rmse: 1652.1
[1056]	valid_0's rmse: 1652.08
[1057]	valid_0's rmse: 1652.05
[1058]	valid_0's rmse: 1652.1
[1059]	valid_0's rmse: 1652.07
[1060]	valid_0's rmse: 1652.04
[1061]	valid_0's rmse: 1652.07
[1062]	valid_0's rmse: 1652.1
[1063]	valid_0's rmse: 1652.13
[1064]	valid_0's rmse: 1651.98
[1065]	valid_0's rmse: 1652
[1066]	valid_0's rmse: 1651.99
[1067]	valid_0's rmse: 1652.07
[1068]	valid_0's rmse: 1652.04
[1069]	valid_0's rmse: 1652.02
[1070]	valid_0's rmse: 1652
[1071]	valid_0's rmse: 1652.01
[1072]	valid_0's rmse: 1651.93
[1073]	valid_0's rmse: 1651.83
[1074]	valid_0's rmse: 1651.83
[1075]	valid_0's rmse: 1651.82
[1076]	valid_0's rmse: 1651.87
[1077]	valid_0's rmse: 1651.85
[1078]	valid_0's rmse: 1651.75
[1079]	valid_0's rmse: 1651.71
[1080]	valid_0's rmse: 1651.76
[1081]	valid_0's rmse: 1651.71
[1082]	valid_0's rmse: 1651.64
[1083]	valid_0's rmse: 1651.57
[1084]	valid_0's rmse: 1651.46
[1085]	valid_0's rmse: 1651.44
[1086]	valid_0's rmse: 1651.39
[1087]	valid_0's rmse: 1651.39
[1088]	valid_0's rmse: 1651.42
[1089]	valid_0's rmse: 1651.34
[1090]	valid_0's rmse: 1651.33
[1091]	valid_0's rmse: 1651.33
[1092]	valid_0's rmse: 1651.37
[1093]	valid_0's rmse: 1651.39
[1094]	valid_0's rmse: 1651.45
[1095]	valid_0's rmse: 1651.39
[1096]	valid_0's rmse: 1651.38
[1097]	valid_0's rmse: 1651.33
[1098]	valid_0's rmse: 1651.34
[1099]	valid_0's rmse: 1651.34
[1100]	valid_0's rmse: 1651.38
[1101]	valid_0's rmse: 1651.39
[1102]	valid_0's rmse: 1651.41
[1103]	valid_0's rmse: 1651.36
[1104]	valid_0's rmse: 1651.3
[1105]	valid_0's rmse: 1651.33
[1106]	valid_0's rmse: 1651.35
[1107]	valid_0's rmse: 1651.38
[1108]	valid_0's rmse: 1651.34
[1109]	valid_0's rmse: 1651.28
[1110]	valid_0's rmse: 1651.28
[1111]	valid_0's rmse: 1651.24
[1112]	valid_0's rmse: 1651.21
[1113]	valid_0's rmse: 1651.18
[1114]	valid_0's rmse: 1651.21
[1115]	valid_0's rmse: 1651.18
[1116]	valid_0's rmse: 1651.08
[1117]	valid_0's rmse: 1651
[1118]	valid_0's rmse: 1651.02
[1119]	valid_0's rmse: 1651.04
[1120]	valid_0's rmse: 1651.02
[1121]	valid_0's rmse: 1651.01
[1122]	valid_0's rmse: 1650.99
[1123]	valid_0's rmse: 1650.99
[1124]	valid_0's rmse: 1650.93
[1125]	valid_0's rmse: 1650.9
[1126]	valid_0's rmse: 1650.78
[1127]	valid_0's rmse: 1650.77
[1128]	valid_0's rmse: 1650.78
[1129]	valid_0's rmse: 1650.71
[1130]	valid_0's rmse: 1650.68
[1131]	valid_0's rmse: 1650.55
[1132]	valid_0's rmse: 1650.58
[1133]	valid_0's rmse: 1650.68
[1134]	valid_0's rmse: 1650.69
[1135]	valid_0's rmse: 1650.67
[1136]	valid_0's rmse: 1650.73
[1137]	valid_0's rmse: 1650.71
[1138]	valid_0's rmse: 1650.64
[1139]	valid_0's rmse: 1650.53
[1140]	valid_0's rmse: 1650.48
[1141]	valid_0's rmse: 1650.52
[1142]	valid_0's rmse: 1650.5
[1143]	valid_0's rmse: 1650.48
[1144]	valid_0's rmse: 1650.51
[1145]	valid_0's rmse: 1650.53
[1146]	valid_0's rmse: 1650.56
[1147]	valid_0's rmse: 1650.4
[1148]	valid_0's rmse: 1650.4
[1149]	valid_0's rmse: 1650.33
[1150]	valid_0's rmse: 1650.21
[1151]	valid_0's rmse: 1650.19
[1152]	valid_0's rmse: 1650.12
[1153]	valid_0's rmse: 1650.02
[1154]	valid_0's rmse: 1649.99
[1155]	valid_0's rmse: 1649.96
[1156]	valid_0's rmse: 1649.84
[1157]	valid_0's rmse: 1649.83
[1158]	valid_0's rmse: 1649.88
[1159]	valid_0's rmse: 1649.87
[1160]	valid_0's rmse: 1649.82
[1161]	valid_0's rmse: 1649.8
[1162]	valid_0's rmse: 1649.85
[1163]	valid_0's rmse: 1649.88
[1164]	valid_0's rmse: 1649.9
[1165]	valid_0's rmse: 1649.81
[1166]	valid_0's rmse: 1649.82
[1167]	valid_0's rmse: 1649.82
[1168]	valid_0's rmse: 1649.82
[1169]	valid_0's rmse: 1649.74
[1170]	valid_0's rmse: 1649.68
[1171]	valid_0's rmse: 1649.56
[1172]	valid_0's rmse: 1649.61
[1173]	valid_0's rmse: 1649.58
[1174]	valid_0's rmse: 1649.6
[1175]	valid_0's rmse: 1649.48
[1176]	valid_0's rmse: 1649.49
[1177]	valid_0's rmse: 1649.47
[1178]	valid_0's rmse: 1649.47
[1179]	valid_0's rmse: 1649.42
[1180]	valid_0's rmse: 1649.37
[1181]	valid_0's rmse: 1649.37
[1182]	valid_0's rmse: 1649.35
[1183]	valid_0's rmse: 1649.33
[1184]	valid_0's rmse: 1649.36
[1185]	valid_0's rmse: 1649.25
[1186]	valid_0's rmse: 1649.21
[1187]	valid_0's rmse: 1649.02
[1188]	valid_0's rmse: 1648.98
[1189]	valid_0's rmse: 1648.91
[1190]	valid_0's rmse: 1648.93
[1191]	valid_0's rmse: 1648.81
[1192]	valid_0's rmse: 1648.7
[1193]	valid_0's rmse: 1648.62
[1194]	valid_0's rmse: 1648.51
[1195]	valid_0's rmse: 1648.48
[1196]	valid_0's rmse: 1648.46
[1197]	valid_0's rmse: 1648.42
[1198]	valid_0's rmse: 1648.43
[1199]	valid_0's rmse: 1648.49
[1200]	valid_0's rmse: 1648.49
[1201]	valid_0's rmse: 1648.41
[1202]	valid_0's rmse: 1648.45
[1203]	valid_0's rmse: 1648.37
[1204]	valid_0's rmse: 1648.24
[1205]	valid_0's rmse: 1648.13
[1206]	valid_0's rmse: 1648.02
[1207]	valid_0's rmse: 1648.07
[1208]	valid_0's rmse: 1648.03
[1209]	valid_0's rmse: 1648.04
[1210]	valid_0's rmse: 1647.93
[1211]	valid_0's rmse: 1647.94
[1212]	valid_0's rmse: 1647.98
[1213]	valid_0's rmse: 1647.98
[1214]	valid_0's rmse: 1647.96
[1215]	valid_0's rmse: 1647.87
[1216]	valid_0's rmse: 1647.9
[1217]	valid_0's rmse: 1647.81
[1218]	valid_0's rmse: 1647.78
[1219]	valid_0's rmse: 1647.78
[1220]	valid_0's rmse: 1647.78
[1221]	valid_0's rmse: 1647.8
[1222]	valid_0's rmse: 1647.79
[1223]	valid_0's rmse: 1647.79
[1224]	valid_0's rmse: 1647.69
[1225]	valid_0's rmse: 1647.64
[1226]	valid_0's rmse: 1647.67
[1227]	valid_0's rmse: 1647.7
[1228]	valid_0's rmse: 1647.73
[1229]	valid_0's rmse: 1647.67
[1230]	valid_0's rmse: 1647.69
[1231]	valid_0's rmse: 1647.52
[1232]	valid_0's rmse: 1647.38
[1233]	valid_0's rmse: 1647.4
[1234]	valid_0's rmse: 1647.4
[1235]	valid_0's rmse: 1647.37
[1236]	valid_0's rmse: 1647.39
[1237]	valid_0's rmse: 1647.37
[1238]	valid_0's rmse: 1647.34
[1239]	valid_0's rmse: 1647.31
[1240]	valid_0's rmse: 1647.29
[1241]	valid_0's rmse: 1647.29
[1242]	valid_0's rmse: 1647.3
[1243]	valid_0's rmse: 1647.31
[1244]	valid_0's rmse: 1647.29
[1245]	valid_0's rmse: 1647.24
[1246]	valid_0's rmse: 1647.24
[1247]	valid_0's rmse: 1647.23
[1248]	valid_0's rmse: 1647.24
[1249]	valid_0's rmse: 1647.26
[1250]	valid_0's rmse: 1647.26
[1251]	valid_0's rmse: 1647.27
[1252]	valid_0's rmse: 1647.29
[1253]	valid_0's rmse: 1647.3
[1254]	valid_0's rmse: 1647.25
[1255]	valid_0's rmse: 1647.24
[1256]	valid_0's rmse: 1647.24
[1257]	valid_0's rmse: 1647.25
[1258]	valid_0's rmse: 1647.24
[1259]	valid_0's rmse: 1647.1
[1260]	valid_0's rmse: 1647.11
[1261]	valid_0's rmse: 1647.11
[1262]	valid_0's rmse: 1647.11
[1263]	valid_0's rmse: 1647.04
[1264]	valid_0's rmse: 1647.08
[1265]	valid_0's rmse: 1647.03
[1266]	valid_0's rmse: 1647.07
[1267]	valid_0's rmse: 1647.07
[1268]	valid_0's rmse: 1647.02
[1269]	valid_0's rmse: 1646.91
[1270]	valid_0's rmse: 1646.97
[1271]	valid_0's rmse: 1647
[1272]	valid_0's rmse: 1647.05
[1273]	valid_0's rmse: 1647.08
[1274]	valid_0's rmse: 1647.03
[1275]	valid_0's rmse: 1647.01
[1276]	valid_0's rmse: 1647.03
[1277]	valid_0's rmse: 1647.05
[1278]	valid_0's rmse: 1647.05
[1279]	valid_0's rmse: 1647.04
[1280]	valid_0's rmse: 1647.03
[1281]	valid_0's rmse: 1646.95
[1282]	valid_0's rmse: 1646.89
[1283]	valid_0's rmse: 1646.84
[1284]	valid_0's rmse: 1646.84
[1285]	valid_0's rmse: 1646.83
[1286]	valid_0's rmse: 1646.81
[1287]	valid_0's rmse: 1646.82
[1288]	valid_0's rmse: 1646.74
[1289]	valid_0's rmse: 1646.76
[1290]	valid_0's rmse: 1646.72
[1291]	valid_0's rmse: 1646.7
[1292]	valid_0's rmse: 1646.62
[1293]	valid_0's rmse: 1646.66
[1294]	valid_0's rmse: 1646.73
[1295]	valid_0's rmse: 1646.71
[1296]	valid_0's rmse: 1646.64
[1297]	valid_0's rmse: 1646.73
[1298]	valid_0's rmse: 1646.71
[1299]	valid_0's rmse: 1646.7
[1300]	valid_0's rmse: 1646.72
[1301]	valid_0's rmse: 1646.73
[1302]	valid_0's rmse: 1646.75
[1303]	valid_0's rmse: 1646.82
[1304]	valid_0's rmse: 1646.83
[1305]	valid_0's rmse: 1646.83
[1306]	valid_0's rmse: 1646.79
[1307]	valid_0's rmse: 1646.76
[1308]	valid_0's rmse: 1646.8
[1309]	valid_0's rmse: 1646.86
[1310]	valid_0's rmse: 1646.89
[1311]	valid_0's rmse: 1646.86
[1312]	valid_0's rmse: 1646.9
[1313]	valid_0's rmse: 1646.95
[1314]	valid_0's rmse: 1646.94
[1315]	valid_0's rmse: 1646.93
[1316]	valid_0's rmse: 1646.92
[1317]	valid_0's rmse: 1646.93
[1318]	valid_0's rmse: 1647
[1319]	valid_0's rmse: 1647.05
[1320]	valid_0's rmse: 1647.03
[1321]	valid_0's rmse: 1647.02
[1322]	valid_0's rmse: 1647.02
[1323]	valid_0's rmse: 1647.04
[1324]	valid_0's rmse: 1647.05
[1325]	valid_0's rmse: 1647.04
[1326]	valid_0's rmse: 1647.05
[1327]	valid_0's rmse: 1647.06
[1328]	valid_0's rmse: 1646.91
[1329]	valid_0's rmse: 1646.89
[1330]	valid_0's rmse: 1646.77
[1331]	valid_0's rmse: 1646.69
[1332]	valid_0's rmse: 1646.73
[1333]	valid_0's rmse: 1646.77
[1334]	valid_0's rmse: 1646.75
[1335]	valid_0's rmse: 1646.76
[1336]	valid_0's rmse: 1646.83
[1337]	valid_0's rmse: 1646.77
[1338]	valid_0's rmse: 1646.75
[1339]	valid_0's rmse: 1646.73
[1340]	valid_0's rmse: 1646.73
[1341]	valid_0's rmse: 1646.72
[1342]	valid_0's rmse: 1646.72
[1343]	valid_0's rmse: 1646.74
[1344]	valid_0's rmse: 1646.72
[1345]	valid_0's rmse: 1646.74
[1346]	valid_0's rmse: 1646.74
[1347]	valid_0's rmse: 1646.73
[1348]	valid_0's rmse: 1646.7
[1349]	valid_0's rmse: 1646.68
[1350]	valid_0's rmse: 1646.73
[1351]	valid_0's rmse: 1646.71
[1352]	valid_0's rmse: 1646.72
[1353]	valid_0's rmse: 1646.72
[1354]	valid_0's rmse: 1646.74
[1355]	valid_0's rmse: 1646.77
[1356]	valid_0's rmse: 1646.72
[1357]	valid_0's rmse: 1646.7
[1358]	valid_0's rmse: 1646.67
[1359]	valid_0's rmse: 1646.59
[1360]	valid_0's rmse: 1646.53
[1361]	valid_0's rmse: 1646.44
[1362]	valid_0's rmse: 1646.34
[1363]	valid_0's rmse: 1646.45
[1364]	valid_0's rmse: 1646.37
[1365]	valid_0's rmse: 1646.36
[1366]	valid_0's rmse: 1646.4
[1367]	valid_0's rmse: 1646.37
[1368]	valid_0's rmse: 1646.2
[1369]	valid_0's rmse: 1646.22
[1370]	valid_0's rmse: 1646.22
[1371]	valid_0's rmse: 1646.24
[1372]	valid_0's rmse: 1646.21
[1373]	valid_0's rmse: 1646.24
[1374]	valid_0's rmse: 1646.24
[1375]	valid_0's rmse: 1646.19
[1376]	valid_0's rmse: 1646.23
[1377]	valid_0's rmse: 1646.27
[1378]	valid_0's rmse: 1646.24
[1379]	valid_0's rmse: 1646.23
[1380]	valid_0's rmse: 1646.23
[1381]	valid_0's rmse: 1646.19
[1382]	valid_0's rmse: 1646.18
[1383]	valid_0's rmse: 1646.13
[1384]	valid_0's rmse: 1646.13
[1385]	valid_0's rmse: 1646.22
[1386]	valid_0's rmse: 1646.19
[1387]	valid_0's rmse: 1646.09
[1388]	valid_0's rmse: 1646.02
[1389]	valid_0's rmse: 1645.98
[1390]	valid_0's rmse: 1645.97
[1391]	valid_0's rmse: 1645.92
[1392]	valid_0's rmse: 1645.89
[1393]	valid_0's rmse: 1645.82
[1394]	valid_0's rmse: 1645.79
[1395]	valid_0's rmse: 1645.75
[1396]	valid_0's rmse: 1645.7
[1397]	valid_0's rmse: 1645.69
[1398]	valid_0's rmse: 1645.65
[1399]	valid_0's rmse: 1645.6
[1400]	valid_0's rmse: 1645.64
[1401]	valid_0's rmse: 1645.63
[1402]	valid_0's rmse: 1645.63
[1403]	valid_0's rmse: 1645.59
[1404]	valid_0's rmse: 1645.57
[1405]	valid_0's rmse: 1645.48
[1406]	valid_0's rmse: 1645.49
[1407]	valid_0's rmse: 1645.47
[1408]	valid_0's rmse: 1645.42
[1409]	valid_0's rmse: 1645.35
[1410]	valid_0's rmse: 1645.37
[1411]	valid_0's rmse: 1645.48
[1412]	valid_0's rmse: 1645.51
[1413]	valid_0's rmse: 1645.48
[1414]	valid_0's rmse: 1645.49
[1415]	valid_0's rmse: 1645.49
[1416]	valid_0's rmse: 1645.43
[1417]	valid_0's rmse: 1645.35
[1418]	valid_0's rmse: 1645.34
[1419]	valid_0's rmse: 1645.25
[1420]	valid_0's rmse: 1645.11
[1421]	valid_0's rmse: 1645.04
[1422]	valid_0's rmse: 1644.99
[1423]	valid_0's rmse: 1645
[1424]	valid_0's rmse: 1645.02
[1425]	valid_0's rmse: 1645.04
[1426]	valid_0's rmse: 1644.93
[1427]	valid_0's rmse: 1644.98
[1428]	valid_0's rmse: 1644.94
[1429]	valid_0's rmse: 1644.87
[1430]	valid_0's rmse: 1644.87
[1431]	valid_0's rmse: 1644.82
[1432]	valid_0's rmse: 1644.8
[1433]	valid_0's rmse: 1644.83
[1434]	valid_0's rmse: 1644.88
[1435]	valid_0's rmse: 1644.82
[1436]	valid_0's rmse: 1644.78
[1437]	valid_0's rmse: 1644.77
[1438]	valid_0's rmse: 1644.69
[1439]	valid_0's rmse: 1644.72
[1440]	valid_0's rmse: 1644.62
[1441]	valid_0's rmse: 1644.57
[1442]	valid_0's rmse: 1644.48
[1443]	valid_0's rmse: 1644.51
[1444]	valid_0's rmse: 1644.61
[1445]	valid_0's rmse: 1644.64
[1446]	valid_0's rmse: 1644.64
[1447]	valid_0's rmse: 1644.6
[1448]	valid_0's rmse: 1644.59
[1449]	valid_0's rmse: 1644.57
[1450]	valid_0's rmse: 1644.51
[1451]	valid_0's rmse: 1644.4
[1452]	valid_0's rmse: 1644.38
[1453]	valid_0's rmse: 1644.35
[1454]	valid_0's rmse: 1644.31
[1455]	valid_0's rmse: 1644.33
[1456]	valid_0's rmse: 1644.36
[1457]	valid_0's rmse: 1644.35
[1458]	valid_0's rmse: 1644.35
[1459]	valid_0's rmse: 1644.34
[1460]	valid_0's rmse: 1644.32
[1461]	valid_0's rmse: 1644.27
[1462]	valid_0's rmse: 1644.27
[1463]	valid_0's rmse: 1644.23
[1464]	valid_0's rmse: 1644.18
[1465]	valid_0's rmse: 1644.29
[1466]	valid_0's rmse: 1644.27
[1467]	valid_0's rmse: 1644.29
[1468]	valid_0's rmse: 1644.19
[1469]	valid_0's rmse: 1644.26
[1470]	valid_0's rmse: 1644.23
[1471]	valid_0's rmse: 1644.22
[1472]	valid_0's rmse: 1644.23
[1473]	valid_0's rmse: 1644.23
[1474]	valid_0's rmse: 1644.25
[1475]	valid_0's rmse: 1644.24
[1476]	valid_0's rmse: 1644.22
[1477]	valid_0's rmse: 1644.21
[1478]	valid_0's rmse: 1644.2
[1479]	valid_0's rmse: 1644.18
[1480]	valid_0's rmse: 1644.18
[1481]	valid_0's rmse: 1644.16
[1482]	valid_0's rmse: 1644.15
[1483]	valid_0's rmse: 1644.13
[1484]	valid_0's rmse: 1644.12
[1485]	valid_0's rmse: 1644.1
[1486]	valid_0's rmse: 1644.11
[1487]	valid_0's rmse: 1644.07
[1488]	valid_0's rmse: 1644.1
[1489]	valid_0's rmse: 1644.08
[1490]	valid_0's rmse: 1644.08
[1491]	valid_0's rmse: 1644.07
[1492]	valid_0's rmse: 1643.93
[1493]	valid_0's rmse: 1643.9
[1494]	valid_0's rmse: 1643.85
[1495]	valid_0's rmse: 1643.84
[1496]	valid_0's rmse: 1643.8
[1497]	valid_0's rmse: 1643.82
[1498]	valid_0's rmse: 1643.8
[1499]	valid_0's rmse: 1643.8
[1500]	valid_0's rmse: 1643.77
[1501]	valid_0's rmse: 1643.81
[1502]	valid_0's rmse: 1643.73
[1503]	valid_0's rmse: 1643.71
[1504]	valid_0's rmse: 1643.56
[1505]	valid_0's rmse: 1643.56
[1506]	valid_0's rmse: 1643.51
[1507]	valid_0's rmse: 1643.45
[1508]	valid_0's rmse: 1643.39
[1509]	valid_0's rmse: 1643.36
[1510]	valid_0's rmse: 1643.32
[1511]	valid_0's rmse: 1643.32
[1512]	valid_0's rmse: 1643.25
[1513]	valid_0's rmse: 1643.25
[1514]	valid_0's rmse: 1643.27
[1515]	valid_0's rmse: 1643.25
[1516]	valid_0's rmse: 1643.25
[1517]	valid_0's rmse: 1643.2
[1518]	valid_0's rmse: 1643.22
[1519]	valid_0's rmse: 1643.22
[1520]	valid_0's rmse: 1643.15
[1521]	valid_0's rmse: 1643.1
[1522]	valid_0's rmse: 1643.14
[1523]	valid_0's rmse: 1643.11
[1524]	valid_0's rmse: 1643.06
[1525]	valid_0's rmse: 1643.03
[1526]	valid_0's rmse: 1642.98
[1527]	valid_0's rmse: 1642.94
[1528]	valid_0's rmse: 1642.94
[1529]	valid_0's rmse: 1642.94
[1530]	valid_0's rmse: 1642.92
[1531]	valid_0's rmse: 1642.84
[1532]	valid_0's rmse: 1642.84
[1533]	valid_0's rmse: 1642.8
[1534]	valid_0's rmse: 1642.75
[1535]	valid_0's rmse: 1642.8
[1536]	valid_0's rmse: 1642.79
[1537]	valid_0's rmse: 1642.75
[1538]	valid_0's rmse: 1642.75
[1539]	valid_0's rmse: 1642.77
[1540]	valid_0's rmse: 1642.84
[1541]	valid_0's rmse: 1642.76
[1542]	valid_0's rmse: 1642.73
[1543]	valid_0's rmse: 1642.65
[1544]	valid_0's rmse: 1642.59
[1545]	valid_0's rmse: 1642.54
[1546]	valid_0's rmse: 1642.55
[1547]	valid_0's rmse: 1642.53
[1548]	valid_0's rmse: 1642.55
[1549]	valid_0's rmse: 1642.5
[1550]	valid_0's rmse: 1642.39
[1551]	valid_0's rmse: 1642.42
[1552]	valid_0's rmse: 1642.33
[1553]	valid_0's rmse: 1642.26
[1554]	valid_0's rmse: 1642.25
[1555]	valid_0's rmse: 1642.25
[1556]	valid_0's rmse: 1642.27
[1557]	valid_0's rmse: 1642.28
[1558]	valid_0's rmse: 1642.2
[1559]	valid_0's rmse: 1642.14
[1560]	valid_0's rmse: 1642.11
[1561]	valid_0's rmse: 1642.09
[1562]	valid_0's rmse: 1642.07
[1563]	valid_0's rmse: 1642.02
[1564]	valid_0's rmse: 1641.96
[1565]	valid_0's rmse: 1641.81
[1566]	valid_0's rmse: 1641.77
[1567]	valid_0's rmse: 1641.74
[1568]	valid_0's rmse: 1641.63
[1569]	valid_0's rmse: 1641.61
[1570]	valid_0's rmse: 1641.54
[1571]	valid_0's rmse: 1641.53
[1572]	valid_0's rmse: 1641.48
[1573]	valid_0's rmse: 1641.51
[1574]	valid_0's rmse: 1641.53
[1575]	valid_0's rmse: 1641.55
[1576]	valid_0's rmse: 1641.55
[1577]	valid_0's rmse: 1641.42
[1578]	valid_0's rmse: 1641.42
[1579]	valid_0's rmse: 1641.38
[1580]	valid_0's rmse: 1641.34
[1581]	valid_0's rmse: 1641.38
[1582]	valid_0's rmse: 1641.41
[1583]	valid_0's rmse: 1641.39
[1584]	valid_0's rmse: 1641.3
[1585]	valid_0's rmse: 1641.31
[1586]	valid_0's rmse: 1641.33
[1587]	valid_0's rmse: 1641.31
[1588]	valid_0's rmse: 1641.3
[1589]	valid_0's rmse: 1641.32
[1590]	valid_0's rmse: 1641.26
[1591]	valid_0's rmse: 1641.26
[1592]	valid_0's rmse: 1641.25
[1593]	valid_0's rmse: 1641.24
[1594]	valid_0's rmse: 1641.23
[1595]	valid_0's rmse: 1641.19
[1596]	valid_0's rmse: 1641.17
[1597]	valid_0's rmse: 1641.19
[1598]	valid_0's rmse: 1641.17
[1599]	valid_0's rmse: 1641.13
[1600]	valid_0's rmse: 1641.05
[1601]	valid_0's rmse: 1641.05
[1602]	valid_0's rmse: 1641.01
[1603]	valid_0's rmse: 1641.05
[1604]	valid_0's rmse: 1641.01
[1605]	valid_0's rmse: 1640.99
[1606]	valid_0's rmse: 1641.02
[1607]	valid_0's rmse: 1641.05
[1608]	valid_0's rmse: 1641
[1609]	valid_0's rmse: 1640.99
[1610]	valid_0's rmse: 1640.87
[1611]	valid_0's rmse: 1640.82
[1612]	valid_0's rmse: 1640.79
[1613]	valid_0's rmse: 1640.8
[1614]	valid_0's rmse: 1640.84
[1615]	valid_0's rmse: 1640.88
[1616]	valid_0's rmse: 1640.84
[1617]	valid_0's rmse: 1640.83
[1618]	valid_0's rmse: 1640.8
[1619]	valid_0's rmse: 1640.84
[1620]	valid_0's rmse: 1640.85
[1621]	valid_0's rmse: 1640.9
[1622]	valid_0's rmse: 1640.85
[1623]	valid_0's rmse: 1640.8
[1624]	valid_0's rmse: 1640.77
[1625]	valid_0's rmse: 1640.81
[1626]	valid_0's rmse: 1640.76
[1627]	valid_0's rmse: 1640.72
[1628]	valid_0's rmse: 1640.69
[1629]	valid_0's rmse: 1640.72
[1630]	valid_0's rmse: 1640.73
[1631]	valid_0's rmse: 1640.71
[1632]	valid_0's rmse: 1640.74
[1633]	valid_0's rmse: 1640.73
[1634]	valid_0's rmse: 1640.71
[1635]	valid_0's rmse: 1640.65
[1636]	valid_0's rmse: 1640.61
[1637]	valid_0's rmse: 1640.59
[1638]	valid_0's rmse: 1640.63
[1639]	valid_0's rmse: 1640.61
[1640]	valid_0's rmse: 1640.56
[1641]	valid_0's rmse: 1640.54
[1642]	valid_0's rmse: 1640.5
[1643]	valid_0's rmse: 1640.58
[1644]	valid_0's rmse: 1640.58
[1645]	valid_0's rmse: 1640.61
[1646]	valid_0's rmse: 1640.59
[1647]	valid_0's rmse: 1640.56
[1648]	valid_0's rmse: 1640.54
[1649]	valid_0's rmse: 1640.54
[1650]	valid_0's rmse: 1640.57
[1651]	valid_0's rmse: 1640.53
[1652]	valid_0's rmse: 1640.5
[1653]	valid_0's rmse: 1640.52
[1654]	valid_0's rmse: 1640.52
[1655]	valid_0's rmse: 1640.46
[1656]	valid_0's rmse: 1640.45
[1657]	valid_0's rmse: 1640.47
[1658]	valid_0's rmse: 1640.38
[1659]	valid_0's rmse: 1640.38
[1660]	valid_0's rmse: 1640.33
[1661]	valid_0's rmse: 1640.27
[1662]	valid_0's rmse: 1640.3
[1663]	valid_0's rmse: 1640.28
[1664]	valid_0's rmse: 1640.34
[1665]	valid_0's rmse: 1640.32
[1666]	valid_0's rmse: 1640.32
[1667]	valid_0's rmse: 1640.31
[1668]	valid_0's rmse: 1640.29
[1669]	valid_0's rmse: 1640.29
[1670]	valid_0's rmse: 1640.29
[1671]	valid_0's rmse: 1640.3
[1672]	valid_0's rmse: 1640.24
[1673]	valid_0's rmse: 1640.21
[1674]	valid_0's rmse: 1640.14
[1675]	valid_0's rmse: 1640.08
[1676]	valid_0's rmse: 1640.11
[1677]	valid_0's rmse: 1640.1
[1678]	valid_0's rmse: 1640.07
[1679]	valid_0's rmse: 1640.03
[1680]	valid_0's rmse: 1640.05
[1681]	valid_0's rmse: 1640.01
[1682]	valid_0's rmse: 1639.97
[1683]	valid_0's rmse: 1639.99
[1684]	valid_0's rmse: 1640.1
[1685]	valid_0's rmse: 1640.13
[1686]	valid_0's rmse: 1640.13
[1687]	valid_0's rmse: 1640.15
[1688]	valid_0's rmse: 1640.13
[1689]	valid_0's rmse: 1640.14
[1690]	valid_0's rmse: 1640.15
[1691]	valid_0's rmse: 1640.16
[1692]	valid_0's rmse: 1640.13
[1693]	valid_0's rmse: 1640.15
[1694]	valid_0's rmse: 1640.15
[1695]	valid_0's rmse: 1640.15
[1696]	valid_0's rmse: 1640.15
[1697]	valid_0's rmse: 1640.17
[1698]	valid_0's rmse: 1640.14
[1699]	valid_0's rmse: 1640.15
[1700]	valid_0's rmse: 1640.11
[1701]	valid_0's rmse: 1640.11
[1702]	valid_0's rmse: 1640.15
[1703]	valid_0's rmse: 1640.14
[1704]	valid_0's rmse: 1640.16
[1705]	valid_0's rmse: 1640.15
[1706]	valid_0's rmse: 1640.2
[1707]	valid_0's rmse: 1640.13
[1708]	valid_0's rmse: 1640.16
[1709]	valid_0's rmse: 1640.11
[1710]	valid_0's rmse: 1640.1
[1711]	valid_0's rmse: 1640.01
[1712]	valid_0's rmse: 1640.06
[1713]	valid_0's rmse: 1640.05
[1714]	valid_0's rmse: 1640.03
[1715]	valid_0's rmse: 1639.9
[1716]	valid_0's rmse: 1639.88
[1717]	valid_0's rmse: 1639.81
[1718]	valid_0's rmse: 1639.86
[1719]	valid_0's rmse: 1639.82
[1720]	valid_0's rmse: 1639.82
[1721]	valid_0's rmse: 1639.71
[1722]	valid_0's rmse: 1639.71
[1723]	valid_0's rmse: 1639.68
[1724]	valid_0's rmse: 1639.68
[1725]	valid_0's rmse: 1639.65
[1726]	valid_0's rmse: 1639.65
[1727]	valid_0's rmse: 1639.54
[1728]	valid_0's rmse: 1639.55
[1729]	valid_0's rmse: 1639.6
[1730]	valid_0's rmse: 1639.57
[1731]	valid_0's rmse: 1639.56
[1732]	valid_0's rmse: 1639.59
[1733]	valid_0's rmse: 1639.54
[1734]	valid_0's rmse: 1639.57
[1735]	valid_0's rmse: 1639.59
[1736]	valid_0's rmse: 1639.59
[1737]	valid_0's rmse: 1639.51
[1738]	valid_0's rmse: 1639.52
[1739]	valid_0's rmse: 1639.55
[1740]	valid_0's rmse: 1639.57
[1741]	valid_0's rmse: 1639.59
[1742]	valid_0's rmse: 1639.59
[1743]	valid_0's rmse: 1639.55
[1744]	valid_0's rmse: 1639.51
[1745]	valid_0's rmse: 1639.52
[1746]	valid_0's rmse: 1639.53
[1747]	valid_0's rmse: 1639.51
[1748]	valid_0's rmse: 1639.56
[1749]	valid_0's rmse: 1639.6
[1750]	valid_0's rmse: 1639.6
[1751]	valid_0's rmse: 1639.58
[1752]	valid_0's rmse: 1639.55
[1753]	valid_0's rmse: 1639.57
[1754]	valid_0's rmse: 1639.55
[1755]	valid_0's rmse: 1639.51
[1756]	valid_0's rmse: 1639.47
[1757]	valid_0's rmse: 1639.54
[1758]	valid_0's rmse: 1639.52
[1759]	valid_0's rmse: 1639.52
[1760]	valid_0's rmse: 1639.41
[1761]	valid_0's rmse: 1639.38
[1762]	valid_0's rmse: 1639.36
[1763]	valid_0's rmse: 1639.3
[1764]	valid_0's rmse: 1639.38
[1765]	valid_0's rmse: 1639.37
[1766]	valid_0's rmse: 1639.35
[1767]	valid_0's rmse: 1639.32
[1768]	valid_0's rmse: 1639.25
[1769]	valid_0's rmse: 1639.25
[1770]	valid_0's rmse: 1639.22
[1771]	valid_0's rmse: 1639.19
[1772]	valid_0's rmse: 1639.17
[1773]	valid_0's rmse: 1639.18
[1774]	valid_0's rmse: 1639.14
[1775]	valid_0's rmse: 1639.17
[1776]	valid_0's rmse: 1639.11
[1777]	valid_0's rmse: 1639
[1778]	valid_0's rmse: 1638.95
[1779]	valid_0's rmse: 1638.9
[1780]	valid_0's rmse: 1638.9
[1781]	valid_0's rmse: 1638.88
[1782]	valid_0's rmse: 1638.86
[1783]	valid_0's rmse: 1638.8
[1784]	valid_0's rmse: 1638.8
[1785]	valid_0's rmse: 1638.8
[1786]	valid_0's rmse: 1638.81
[1787]	valid_0's rmse: 1638.78
[1788]	valid_0's rmse: 1638.84
[1789]	valid_0's rmse: 1638.84
[1790]	valid_0's rmse: 1638.86
[1791]	valid_0's rmse: 1638.86
[1792]	valid_0's rmse: 1638.86
[1793]	valid_0's rmse: 1638.81
[1794]	valid_0's rmse: 1638.81
[1795]	valid_0's rmse: 1638.81
[1796]	valid_0's rmse: 1638.77
[1797]	valid_0's rmse: 1638.77
[1798]	valid_0's rmse: 1638.74
[1799]	valid_0's rmse: 1638.71
[1800]	valid_0's rmse: 1638.7
[1801]	valid_0's rmse: 1638.64
[1802]	valid_0's rmse: 1638.67
[1803]	valid_0's rmse: 1638.64
[1804]	valid_0's rmse: 1638.62
[1805]	valid_0's rmse: 1638.63
[1806]	valid_0's rmse: 1638.66
[1807]	valid_0's rmse: 1638.65
[1808]	valid_0's rmse: 1638.65
[1809]	valid_0's rmse: 1638.67
[1810]	valid_0's rmse: 1638.65
[1811]	valid_0's rmse: 1638.65
[1812]	valid_0's rmse: 1638.59
[1813]	valid_0's rmse: 1638.63
[1814]	valid_0's rmse: 1638.62
[1815]	valid_0's rmse: 1638.58
[1816]	valid_0's rmse: 1638.6
[1817]	valid_0's rmse: 1638.61
[1818]	valid_0's rmse: 1638.58
[1819]	valid_0's rmse: 1638.54
[1820]	valid_0's rmse: 1638.46
[1821]	valid_0's rmse: 1638.51
[1822]	valid_0's rmse: 1638.51
[1823]	valid_0's rmse: 1638.56
[1824]	valid_0's rmse: 1638.52
[1825]	valid_0's rmse: 1638.48
[1826]	valid_0's rmse: 1638.43
[1827]	valid_0's rmse: 1638.44
[1828]	valid_0's rmse: 1638.44
[1829]	valid_0's rmse: 1638.44
[1830]	valid_0's rmse: 1638.41
[1831]	valid_0's rmse: 1638.34
[1832]	valid_0's rmse: 1638.34
[1833]	valid_0's rmse: 1638.34
[1834]	valid_0's rmse: 1638.33
[1835]	valid_0's rmse: 1638.35
[1836]	valid_0's rmse: 1638.36
[1837]	valid_0's rmse: 1638.36
[1838]	valid_0's rmse: 1638.32
[1839]	valid_0's rmse: 1638.31
[1840]	valid_0's rmse: 1638.32
[1841]	valid_0's rmse: 1638.32
[1842]	valid_0's rmse: 1638.34
[1843]	valid_0's rmse: 1638.38
[1844]	valid_0's rmse: 1638.4
[1845]	valid_0's rmse: 1638.41
[1846]	valid_0's rmse: 1638.4
[1847]	valid_0's rmse: 1638.43
[1848]	valid_0's rmse: 1638.43
[1849]	valid_0's rmse: 1638.45
[1850]	valid_0's rmse: 1638.47
[1851]	valid_0's rmse: 1638.42
[1852]	valid_0's rmse: 1638.44
[1853]	valid_0's rmse: 1638.41
[1854]	valid_0's rmse: 1638.4
[1855]	valid_0's rmse: 1638.44
[1856]	valid_0's rmse: 1638.46
[1857]	valid_0's rmse: 1638.46
[1858]	valid_0's rmse: 1638.46
[1859]	valid_0's rmse: 1638.46
[1860]	valid_0's rmse: 1638.51
[1861]	valid_0's rmse: 1638.46
[1862]	valid_0's rmse: 1638.45
[1863]	valid_0's rmse: 1638.45
[1864]	valid_0's rmse: 1638.38
[1865]	valid_0's rmse: 1638.33
[1866]	valid_0's rmse: 1638.32
[1867]	valid_0's rmse: 1638.31
[1868]	valid_0's rmse: 1638.28
[1869]	valid_0's rmse: 1638.3
[1870]	valid_0's rmse: 1638.2
[1871]	valid_0's rmse: 1638.22
[1872]	valid_0's rmse: 1638.22
[1873]	valid_0's rmse: 1638.22
[1874]	valid_0's rmse: 1638.13
[1875]	valid_0's rmse: 1638.09
[1876]	valid_0's rmse: 1638.06
[1877]	valid_0's rmse: 1637.98
[1878]	valid_0's rmse: 1638
[1879]	valid_0's rmse: 1637.97
[1880]	valid_0's rmse: 1637.98
[1881]	valid_0's rmse: 1637.99
[1882]	valid_0's rmse: 1637.97
[1883]	valid_0's rmse: 1637.96
[1884]	valid_0's rmse: 1637.95
[1885]	valid_0's rmse: 1637.94
[1886]	valid_0's rmse: 1637.94
[1887]	valid_0's rmse: 1637.96
[1888]	valid_0's rmse: 1637.91
[1889]	valid_0's rmse: 1637.92
[1890]	valid_0's rmse: 1637.89
[1891]	valid_0's rmse: 1637.96
[1892]	valid_0's rmse: 1637.98
[1893]	valid_0's rmse: 1638.02
[1894]	valid_0's rmse: 1637.97
[1895]	valid_0's rmse: 1637.97
[1896]	valid_0's rmse: 1637.99
[1897]	valid_0's rmse: 1638
[1898]	valid_0's rmse: 1638
[1899]	valid_0's rmse: 1638.03
[1900]	valid_0's rmse: 1637.99
[1901]	valid_0's rmse: 1638.02
[1902]	valid_0's rmse: 1638.03
[1903]	valid_0's rmse: 1638.02
[1904]	valid_0's rmse: 1638.05
[1905]	valid_0's rmse: 1638.08
[1906]	valid_0's rmse: 1638.07
[1907]	valid_0's rmse: 1638.12
[1908]	valid_0's rmse: 1638.12
[1909]	valid_0's rmse: 1638.13
[1910]	valid_0's rmse: 1638.16
[1911]	valid_0's rmse: 1638.18
[1912]	valid_0's rmse: 1638.17
[1913]	valid_0's rmse: 1638.13
[1914]	valid_0's rmse: 1638.12
[1915]	valid_0's rmse: 1638.08
[1916]	valid_0's rmse: 1638.04
[1917]	valid_0's rmse: 1638.06
[1918]	valid_0's rmse: 1638.09
[1919]	valid_0's rmse: 1638.11
[1920]	valid_0's rmse: 1638.13
[1921]	valid_0's rmse: 1638.08
[1922]	valid_0's rmse: 1638.13
[1923]	valid_0's rmse: 1638.11
[1924]	valid_0's rmse: 1638.06
[1925]	valid_0's rmse: 1638.04
[1926]	valid_0's rmse: 1638.03
[1927]	valid_0's rmse: 1638.03
[1928]	valid_0's rmse: 1638
[1929]	valid_0's rmse: 1638.03
[1930]	valid_0's rmse: 1638.06
[1931]	valid_0's rmse: 1638.08
[1932]	valid_0's rmse: 1638.08
[1933]	valid_0's rmse: 1637.95
[1934]	valid_0's rmse: 1637.92
[1935]	valid_0's rmse: 1637.86
[1936]	valid_0's rmse: 1637.76
[1937]	valid_0's rmse: 1637.73
[1938]	valid_0's rmse: 1637.71
[1939]	valid_0's rmse: 1637.74
[1940]	valid_0's rmse: 1637.78
[1941]	valid_0's rmse: 1637.77
[1942]	valid_0's rmse: 1637.75
[1943]	valid_0's rmse: 1637.74
[1944]	valid_0's rmse: 1637.77
[1945]	valid_0's rmse: 1637.76
[1946]	valid_0's rmse: 1637.76
[1947]	valid_0's rmse: 1637.79
[1948]	valid_0's rmse: 1637.79
[1949]	valid_0's rmse: 1637.78
[1950]	valid_0's rmse: 1637.78
[1951]	valid_0's rmse: 1637.74
[1952]	valid_0's rmse: 1637.75
[1953]	valid_0's rmse: 1637.72
[1954]	valid_0's rmse: 1637.76
[1955]	valid_0's rmse: 1637.77
[1956]	valid_0's rmse: 1637.8
[1957]	valid_0's rmse: 1637.81
[1958]	valid_0's rmse: 1637.84
[1959]	valid_0's rmse: 1637.81
[1960]	valid_0's rmse: 1637.79
[1961]	valid_0's rmse: 1637.82
[1962]	valid_0's rmse: 1637.84
[1963]	valid_0's rmse: 1637.87
[1964]	valid_0's rmse: 1637.83
[1965]	valid_0's rmse: 1637.83
[1966]	valid_0's rmse: 1637.79
[1967]	valid_0's rmse: 1637.79
[1968]	valid_0's rmse: 1637.83
[1969]	valid_0's rmse: 1637.77
[1970]	valid_0's rmse: 1637.76
[1971]	valid_0's rmse: 1637.71
[1972]	valid_0's rmse: 1637.69
[1973]	valid_0's rmse: 1637.73
[1974]	valid_0's rmse: 1637.72
[1975]	valid_0's rmse: 1637.75
[1976]	valid_0's rmse: 1637.7
[1977]	valid_0's rmse: 1637.68
[1978]	valid_0's rmse: 1637.71
[1979]	valid_0's rmse: 1637.68
[1980]	valid_0's rmse: 1637.62
[1981]	valid_0's rmse: 1637.58
[1982]	valid_0's rmse: 1637.6
[1983]	valid_0's rmse: 1637.54
[1984]	valid_0's rmse: 1637.48
[1985]	valid_0's rmse: 1637.46
[1986]	valid_0's rmse: 1637.44
[1987]	valid_0's rmse: 1637.41
[1988]	valid_0's rmse: 1637.39
[1989]	valid_0's rmse: 1637.46
[1990]	valid_0's rmse: 1637.44
[1991]	valid_0's rmse: 1637.43
[1992]	valid_0's rmse: 1637.39
[1993]	valid_0's rmse: 1637.39
[1994]	valid_0's rmse: 1637.38
[1995]	valid_0's rmse: 1637.27
[1996]	valid_0's rmse: 1637.36
[1997]	valid_0's rmse: 1637.4
[1998]	valid_0's rmse: 1637.4
[1999]	valid_0's rmse: 1637.38
[2000]	valid_0's rmse: 1637.29
[2001]	valid_0's rmse: 1637.32
[2002]	valid_0's rmse: 1637.33
[2003]	valid_0's rmse: 1637.32
[2004]	valid_0's rmse: 1637.32
[2005]	valid_0's rmse: 1637.32
[2006]	valid_0's rmse: 1637.31
[2007]	valid_0's rmse: 1637.31
[2008]	valid_0's rmse: 1637.35
[2009]	valid_0's rmse: 1637.41
[2010]	valid_0's rmse: 1637.47
[2011]	valid_0's rmse: 1637.43
[2012]	valid_0's rmse: 1637.45
[2013]	valid_0's rmse: 1637.45
[2014]	valid_0's rmse: 1637.41
[2015]	valid_0's rmse: 1637.41
[2016]	valid_0's rmse: 1637.42
[2017]	valid_0's rmse: 1637.42
[2018]	valid_0's rmse: 1637.42
[2019]	valid_0's rmse: 1637.36
[2020]	valid_0's rmse: 1637.34
[2021]	valid_0's rmse: 1637.32
[2022]	valid_0's rmse: 1637.32
[2023]	valid_0's rmse: 1637.35
[2024]	valid_0's rmse: 1637.41
[2025]	valid_0's rmse: 1637.41
[2026]	valid_0's rmse: 1637.41
[2027]	valid_0's rmse: 1637.45
[2028]	valid_0's rmse: 1637.44
[2029]	valid_0's rmse: 1637.42
[2030]	valid_0's rmse: 1637.41
[2031]	valid_0's rmse: 1637.43
[2032]	valid_0's rmse: 1637.45
[2033]	valid_0's rmse: 1637.43
[2034]	valid_0's rmse: 1637.45
[2035]	valid_0's rmse: 1637.43
[2036]	valid_0's rmse: 1637.4
[2037]	valid_0's rmse: 1637.39
[2038]	valid_0's rmse: 1637.41
[2039]	valid_0's rmse: 1637.41
[2040]	valid_0's rmse: 1637.41
[2041]	valid_0's rmse: 1637.43
[2042]	valid_0's rmse: 1637.4
[2043]	valid_0's rmse: 1637.37
[2044]	valid_0's rmse: 1637.32
[2045]	valid_0's rmse: 1637.3
[2046]	valid_0's rmse: 1637.25
[2047]	valid_0's rmse: 1637.23
[2048]	valid_0's rmse: 1637.24
[2049]	valid_0's rmse: 1637.25
[2050]	valid_0's rmse: 1637.26
[2051]	valid_0's rmse: 1637.23
[2052]	valid_0's rmse: 1637.21
[2053]	valid_0's rmse: 1637.3
[2054]	valid_0's rmse: 1637.31
[2055]	valid_0's rmse: 1637.25
[2056]	valid_0's rmse: 1637.21
[2057]	valid_0's rmse: 1637.18
[2058]	valid_0's rmse: 1637.18
[2059]	valid_0's rmse: 1637.2
[2060]	valid_0's rmse: 1637.19
[2061]	valid_0's rmse: 1637.2
[2062]	valid_0's rmse: 1637.18
[2063]	valid_0's rmse: 1637.17
[2064]	valid_0's rmse: 1637.09
[2065]	valid_0's rmse: 1637.08
[2066]	valid_0's rmse: 1637.07
[2067]	valid_0's rmse: 1637.06
[2068]	valid_0's rmse: 1636.98
[2069]	valid_0's rmse: 1636.98
[2070]	valid_0's rmse: 1636.94
[2071]	valid_0's rmse: 1636.98
[2072]	valid_0's rmse: 1636.91
[2073]	valid_0's rmse: 1636.89
[2074]	valid_0's rmse: 1636.88
[2075]	valid_0's rmse: 1636.92
[2076]	valid_0's rmse: 1636.84
[2077]	valid_0's rmse: 1636.83
[2078]	valid_0's rmse: 1636.89
[2079]	valid_0's rmse: 1636.9
[2080]	valid_0's rmse: 1636.93
[2081]	valid_0's rmse: 1636.85
[2082]	valid_0's rmse: 1636.87
[2083]	valid_0's rmse: 1636.86
[2084]	valid_0's rmse: 1636.79
[2085]	valid_0's rmse: 1636.78
[2086]	valid_0's rmse: 1636.73
[2087]	valid_0's rmse: 1636.64
[2088]	valid_0's rmse: 1636.66
[2089]	valid_0's rmse: 1636.72
[2090]	valid_0's rmse: 1636.76
[2091]	valid_0's rmse: 1636.78
[2092]	valid_0's rmse: 1636.79
[2093]	valid_0's rmse: 1636.82
[2094]	valid_0's rmse: 1636.84
[2095]	valid_0's rmse: 1636.85
[2096]	valid_0's rmse: 1636.82
[2097]	valid_0's rmse: 1636.84
[2098]	valid_0's rmse: 1636.85
[2099]	valid_0's rmse: 1636.85
[2100]	valid_0's rmse: 1636.86
[2101]	valid_0's rmse: 1636.86
[2102]	valid_0's rmse: 1636.81
[2103]	valid_0's rmse: 1636.81
[2104]	valid_0's rmse: 1636.87
[2105]	valid_0's rmse: 1636.84
[2106]	valid_0's rmse: 1636.85
[2107]	valid_0's rmse: 1636.85
[2108]	valid_0's rmse: 1636.85
[2109]	valid_0's rmse: 1636.85
[2110]	valid_0's rmse: 1636.81
[2111]	valid_0's rmse: 1636.8
[2112]	valid_0's rmse: 1636.83
[2113]	valid_0's rmse: 1636.82
[2114]	valid_0's rmse: 1636.85
[2115]	valid_0's rmse: 1636.8
[2116]	valid_0's rmse: 1636.79
[2117]	valid_0's rmse: 1636.76
[2118]	valid_0's rmse: 1636.74
[2119]	valid_0's rmse: 1636.75
[2120]	valid_0's rmse: 1636.78
[2121]	valid_0's rmse: 1636.86
[2122]	valid_0's rmse: 1636.83
[2123]	valid_0's rmse: 1636.87
[2124]	valid_0's rmse: 1636.85
[2125]	valid_0's rmse: 1636.86
[2126]	valid_0's rmse: 1636.78
[2127]	valid_0's rmse: 1636.77
[2128]	valid_0's rmse: 1636.69
[2129]	valid_0's rmse: 1636.67
[2130]	valid_0's rmse: 1636.65
[2131]	valid_0's rmse: 1636.69
[2132]	valid_0's rmse: 1636.61
[2133]	valid_0's rmse: 1636.54
[2134]	valid_0's rmse: 1636.57
[2135]	valid_0's rmse: 1636.58
[2136]	valid_0's rmse: 1636.47
[2137]	valid_0's rmse: 1636.46
[2138]	valid_0's rmse: 1636.45
[2139]	valid_0's rmse: 1636.46
[2140]	valid_0's rmse: 1636.49
[2141]	valid_0's rmse: 1636.5
[2142]	valid_0's rmse: 1636.49
[2143]	valid_0's rmse: 1636.51
[2144]	valid_0's rmse: 1636.46
[2145]	valid_0's rmse: 1636.46
[2146]	valid_0's rmse: 1636.45
[2147]	valid_0's rmse: 1636.53
[2148]	valid_0's rmse: 1636.58
[2149]	valid_0's rmse: 1636.59
[2150]	valid_0's rmse: 1636.61
[2151]	valid_0's rmse: 1636.6
[2152]	valid_0's rmse: 1636.6
[2153]	valid_0's rmse: 1636.59
[2154]	valid_0's rmse: 1636.57
[2155]	valid_0's rmse: 1636.59
[2156]	valid_0's rmse: 1636.58
[2157]	valid_0's rmse: 1636.57
[2158]	valid_0's rmse: 1636.55
[2159]	valid_0's rmse: 1636.6
[2160]	valid_0's rmse: 1636.59
[2161]	valid_0's rmse: 1636.58
[2162]	valid_0's rmse: 1636.57
[2163]	valid_0's rmse: 1636.57
[2164]	valid_0's rmse: 1636.57
[2165]	valid_0's rmse: 1636.53
[2166]	valid_0's rmse: 1636.51
[2167]	valid_0's rmse: 1636.51
[2168]	valid_0's rmse: 1636.46
[2169]	valid_0's rmse: 1636.46
[2170]	valid_0's rmse: 1636.46
[2171]	valid_0's rmse: 1636.46
[2172]	valid_0's rmse: 1636.48
[2173]	valid_0's rmse: 1636.49
[2174]	valid_0's rmse: 1636.48
[2175]	valid_0's rmse: 1636.47
[2176]	valid_0's rmse: 1636.44
[2177]	valid_0's rmse: 1636.45
[2178]	valid_0's rmse: 1636.37
[2179]	valid_0's rmse: 1636.36
[2180]	valid_0's rmse: 1636.38
[2181]	valid_0's rmse: 1636.39
[2182]	valid_0's rmse: 1636.39
[2183]	valid_0's rmse: 1636.41
[2184]	valid_0's rmse: 1636.42
[2185]	valid_0's rmse: 1636.44
[2186]	valid_0's rmse: 1636.45
[2187]	valid_0's rmse: 1636.47
[2188]	valid_0's rmse: 1636.52
[2189]	valid_0's rmse: 1636.63
[2190]	valid_0's rmse: 1636.63
[2191]	valid_0's rmse: 1636.64
[2192]	valid_0's rmse: 1636.62
[2193]	valid_0's rmse: 1636.63
[2194]	valid_0's rmse: 1636.62
[2195]	valid_0's rmse: 1636.67
[2196]	valid_0's rmse: 1636.63
[2197]	valid_0's rmse: 1636.63
[2198]	valid_0's rmse: 1636.59
[2199]	valid_0's rmse: 1636.59
[2200]	valid_0's rmse: 1636.58
[2201]	valid_0's rmse: 1636.62
[2202]	valid_0's rmse: 1636.58
[2203]	valid_0's rmse: 1636.57
[2204]	valid_0's rmse: 1636.59
[2205]	valid_0's rmse: 1636.62
[2206]	valid_0's rmse: 1636.65
[2207]	valid_0's rmse: 1636.6
[2208]	valid_0's rmse: 1636.6
[2209]	valid_0's rmse: 1636.61
[2210]	valid_0's rmse: 1636.57
[2211]	valid_0's rmse: 1636.57
[2212]	valid_0's rmse: 1636.59
[2213]	valid_0's rmse: 1636.61
[2214]	valid_0's rmse: 1636.63
[2215]	valid_0's rmse: 1636.67
[2216]	valid_0's rmse: 1636.68
[2217]	valid_0's rmse: 1636.71
[2218]	valid_0's rmse: 1636.7
[2219]	valid_0's rmse: 1636.74
[2220]	valid_0's rmse: 1636.8
[2221]	valid_0's rmse: 1636.8
[2222]	valid_0's rmse: 1636.81
[2223]	valid_0's rmse: 1636.81
[2224]	valid_0's rmse: 1636.79
[2225]	valid_0's rmse: 1636.81
[2226]	valid_0's rmse: 1636.81
[2227]	valid_0's rmse: 1636.81
[2228]	valid_0's rmse: 1636.82
[2229]	valid_0's rmse: 1636.81
[2230]	valid_0's rmse: 1636.82
[2231]	valid_0's rmse: 1636.84
[2232]	valid_0's rmse: 1636.8
[2233]	valid_0's rmse: 1636.74
[2234]	valid_0's rmse: 1636.77
[2235]	valid_0's rmse: 1636.7
[2236]	valid_0's rmse: 1636.72
[2237]	valid_0's rmse: 1636.67
[2238]	valid_0's rmse: 1636.68
[2239]	valid_0's rmse: 1636.71
[2240]	valid_0's rmse: 1636.7
[2241]	valid_0's rmse: 1636.7
[2242]	valid_0's rmse: 1636.7
[2243]	valid_0's rmse: 1636.7
[2244]	valid_0's rmse: 1636.7
[2245]	valid_0's rmse: 1636.72
[2246]	valid_0's rmse: 1636.64
[2247]	valid_0's rmse: 1636.66
[2248]	valid_0's rmse: 1636.69
[2249]	valid_0's rmse: 1636.74
[2250]	valid_0's rmse: 1636.71
[2251]	valid_0's rmse: 1636.68
[2252]	valid_0's rmse: 1636.7
[2253]	valid_0's rmse: 1636.71
[2254]	valid_0's rmse: 1636.72
[2255]	valid_0's rmse: 1636.79
[2256]	valid_0's rmse: 1636.76
[2257]	valid_0's rmse: 1636.77
[2258]	valid_0's rmse: 1636.71
[2259]	valid_0's rmse: 1636.72
[2260]	valid_0's rmse: 1636.71
[2261]	valid_0's rmse: 1636.76
[2262]	valid_0's rmse: 1636.84
[2263]	valid_0's rmse: 1636.87
[2264]	valid_0's rmse: 1636.86
[2265]	valid_0's rmse: 1636.85
[2266]	valid_0's rmse: 1636.85
[2267]	valid_0's rmse: 1636.82
[2268]	valid_0's rmse: 1636.88
[2269]	valid_0's rmse: 1636.89
[2270]	valid_0's rmse: 1636.92
[2271]	valid_0's rmse: 1636.89
[2272]	valid_0's rmse: 1636.9
[2273]	valid_0's rmse: 1636.9
[2274]	valid_0's rmse: 1636.9
[2275]	valid_0's rmse: 1636.91
[2276]	valid_0's rmse: 1636.91
[2277]	valid_0's rmse: 1636.91
[2278]	valid_0's rmse: 1636.9
[2279]	valid_0's rmse: 1636.9
[2280]	valid_0's rmse: 1636.86
[2281]	valid_0's rmse: 1636.8
[2282]	valid_0's rmse: 1636.87
[2283]	valid_0's rmse: 1636.88
[2284]	valid_0's rmse: 1636.88
[2285]	valid_0's rmse: 1636.88
[2286]	valid_0's rmse: 1636.85
[2287]	valid_0's rmse: 1636.84
[2288]	valid_0's rmse: 1636.78
[2289]	valid_0's rmse: 1636.78
[2290]	valid_0's rmse: 1636.74
[2291]	valid_0's rmse: 1636.79
[2292]	valid_0's rmse: 1636.78
[2293]	valid_0's rmse: 1636.8
[2294]	valid_0's rmse: 1636.77
[2295]	valid_0's rmse: 1636.79
[2296]	valid_0's rmse: 1636.73
[2297]	valid_0's rmse: 1636.76
[2298]	valid_0's rmse: 1636.76
[2299]	valid_0's rmse: 1636.77
[2300]	valid_0's rmse: 1636.78
[2301]	valid_0's rmse: 1636.8
[2302]	valid_0's rmse: 1636.82
[2303]	valid_0's rmse: 1636.78
[2304]	valid_0's rmse: 1636.78
[2305]	valid_0's rmse: 1636.8
[2306]	valid_0's rmse: 1636.79
[2307]	valid_0's rmse: 1636.71
[2308]	valid_0's rmse: 1636.7
[2309]	valid_0's rmse: 1636.69
[2310]	valid_0's rmse: 1636.72
[2311]	valid_0's rmse: 1636.7
[2312]	valid_0's rmse: 1636.7
[2313]	valid_0's rmse: 1636.7
[2314]	valid_0's rmse: 1636.7
[2315]	valid_0's rmse: 1636.72
[2316]	valid_0's rmse: 1636.76
[2317]	valid_0's rmse: 1636.71
[2318]	valid_0's rmse: 1636.7
[2319]	valid_0's rmse: 1636.72
[2320]	valid_0's rmse: 1636.72
[2321]	valid_0's rmse: 1636.67
[2322]	valid_0's rmse: 1636.67
[2323]	valid_0's rmse: 1636.67
[2324]	valid_0's rmse: 1636.69
[2325]	valid_0's rmse: 1636.67
[2326]	valid_0's rmse: 1636.7
[2327]	valid_0's rmse: 1636.72
[2328]	valid_0's rmse: 1636.68
[2329]	valid_0's rmse: 1636.69
[2330]	valid_0's rmse: 1636.7
[2331]	valid_0's rmse: 1636.7
[2332]	valid_0's rmse: 1636.67
[2333]	valid_0's rmse: 1636.67
[2334]	valid_0's rmse: 1636.7
[2335]	valid_0's rmse: 1636.68
[2336]	valid_0's rmse: 1636.71
[2337]	valid_0's rmse: 1636.72
[2338]	valid_0's rmse: 1636.73
[2339]	valid_0's rmse: 1636.69
[2340]	valid_0's rmse: 1636.66
[2341]	valid_0's rmse: 1636.67
[2342]	valid_0's rmse: 1636.65
[2343]	valid_0's rmse: 1636.66
[2344]	valid_0's rmse: 1636.65
[2345]	valid_0's rmse: 1636.62
[2346]	valid_0's rmse: 1636.61
[2347]	valid_0's rmse: 1636.65
[2348]	valid_0's rmse: 1636.67
[2349]	valid_0's rmse: 1636.62
[2350]	valid_0's rmse: 1636.65
[2351]	valid_0's rmse: 1636.7
[2352]	valid_0's rmse: 1636.71
[2353]	valid_0's rmse: 1636.75
[2354]	valid_0's rmse: 1636.69
[2355]	valid_0's rmse: 1636.65
[2356]	valid_0's rmse: 1636.64
[2357]	valid_0's rmse: 1636.67
[2358]	valid_0's rmse: 1636.67
[2359]	valid_0's rmse: 1636.71
[2360]	valid_0's rmse: 1636.74
[2361]	valid_0's rmse: 1636.74
[2362]	valid_0's rmse: 1636.74
[2363]	valid_0's rmse: 1636.72
[2364]	valid_0's rmse: 1636.69
[2365]	valid_0's rmse: 1636.72
[2366]	valid_0's rmse: 1636.71
[2367]	valid_0's rmse: 1636.72
[2368]	valid_0's rmse: 1636.69
[2369]	valid_0's rmse: 1636.67
[2370]	valid_0's rmse: 1636.65
[2371]	valid_0's rmse: 1636.66
[2372]	valid_0's rmse: 1636.59
[2373]	valid_0's rmse: 1636.55
[2374]	valid_0's rmse: 1636.59
[2375]	valid_0's rmse: 1636.57
[2376]	valid_0's rmse: 1636.54
[2377]	valid_0's rmse: 1636.57
[2378]	valid_0's rmse: 1636.57
[2379]	valid_0's rmse: 1636.54
[2380]	valid_0's rmse: 1636.52
[2381]	valid_0's rmse: 1636.49
[2382]	valid_0's rmse: 1636.46
[2383]	valid_0's rmse: 1636.44
[2384]	valid_0's rmse: 1636.39
[2385]	valid_0's rmse: 1636.4
[2386]	valid_0's rmse: 1636.38
[2387]	valid_0's rmse: 1636.42
[2388]	valid_0's rmse: 1636.41
[2389]	valid_0's rmse: 1636.37
[2390]	valid_0's rmse: 1636.37
[2391]	valid_0's rmse: 1636.35
[2392]	valid_0's rmse: 1636.34
[2393]	valid_0's rmse: 1636.29
[2394]	valid_0's rmse: 1636.28
[2395]	valid_0's rmse: 1636.25
[2396]	valid_0's rmse: 1636.24
[2397]	valid_0's rmse: 1636.26
[2398]	valid_0's rmse: 1636.28
[2399]	valid_0's rmse: 1636.25
[2400]	valid_0's rmse: 1636.24
[2401]	valid_0's rmse: 1636.27
[2402]	valid_0's rmse: 1636.3
[2403]	valid_0's rmse: 1636.24
[2404]	valid_0's rmse: 1636.26
[2405]	valid_0's rmse: 1636.2
[2406]	valid_0's rmse: 1636.24
[2407]	valid_0's rmse: 1636.24
[2408]	valid_0's rmse: 1636.25
[2409]	valid_0's rmse: 1636.22
[2410]	valid_0's rmse: 1636.2
[2411]	valid_0's rmse: 1636.18
[2412]	valid_0's rmse: 1636.19
[2413]	valid_0's rmse: 1636.2
[2414]	valid_0's rmse: 1636.23
[2415]	valid_0's rmse: 1636.24
[2416]	valid_0's rmse: 1636.21
[2417]	valid_0's rmse: 1636.2
[2418]	valid_0's rmse: 1636.21
[2419]	valid_0's rmse: 1636.23
[2420]	valid_0's rmse: 1636.25
[2421]	valid_0's rmse: 1636.23
[2422]	valid_0's rmse: 1636.25
[2423]	valid_0's rmse: 1636.28
[2424]	valid_0's rmse: 1636.24
[2425]	valid_0's rmse: 1636.25
[2426]	valid_0's rmse: 1636.25
[2427]	valid_0's rmse: 1636.22
[2428]	valid_0's rmse: 1636.22
[2429]	valid_0's rmse: 1636.22
[2430]	valid_0's rmse: 1636.19
[2431]	valid_0's rmse: 1636.22
[2432]	valid_0's rmse: 1636.19
[2433]	valid_0's rmse: 1636.14
[2434]	valid_0's rmse: 1636.14
[2435]	valid_0's rmse: 1636.1
[2436]	valid_0's rmse: 1636.09
[2437]	valid_0's rmse: 1636.05
[2438]	valid_0's rmse: 1636.08
[2439]	valid_0's rmse: 1636.09
[2440]	valid_0's rmse: 1636.07
[2441]	valid_0's rmse: 1636.07
[2442]	valid_0's rmse: 1636.08
[2443]	valid_0's rmse: 1636.12
[2444]	valid_0's rmse: 1636.1
[2445]	valid_0's rmse: 1636.11
[2446]	valid_0's rmse: 1636.1
[2447]	valid_0's rmse: 1636.09
[2448]	valid_0's rmse: 1636.09
[2449]	valid_0's rmse: 1636.09
[2450]	valid_0's rmse: 1636.09
[2451]	valid_0's rmse: 1636.07
[2452]	valid_0's rmse: 1636.04
[2453]	valid_0's rmse: 1636.04
[2454]	valid_0's rmse: 1636.08
[2455]	valid_0's rmse: 1636.04
[2456]	valid_0's rmse: 1636.03
[2457]	valid_0's rmse: 1636.02
[2458]	valid_0's rmse: 1636.06
[2459]	valid_0's rmse: 1636.07
[2460]	valid_0's rmse: 1636.09
[2461]	valid_0's rmse: 1636.05
[2462]	valid_0's rmse: 1636.06
[2463]	valid_0's rmse: 1636.03
[2464]	valid_0's rmse: 1635.99
[2465]	valid_0's rmse: 1636.03
[2466]	valid_0's rmse: 1636.06
[2467]	valid_0's rmse: 1636.02
[2468]	valid_0's rmse: 1636
[2469]	valid_0's rmse: 1636
[2470]	valid_0's rmse: 1635.97
[2471]	valid_0's rmse: 1635.98
[2472]	valid_0's rmse: 1635.96
[2473]	valid_0's rmse: 1636
[2474]	valid_0's rmse: 1635.96
[2475]	valid_0's rmse: 1635.96
[2476]	valid_0's rmse: 1635.97
[2477]	valid_0's rmse: 1635.98
[2478]	valid_0's rmse: 1636.01
[2479]	valid_0's rmse: 1636.02
[2480]	valid_0's rmse: 1636.02
[2481]	valid_0's rmse: 1636.03
[2482]	valid_0's rmse: 1636.03
[2483]	valid_0's rmse: 1636.01
[2484]	valid_0's rmse: 1635.97
[2485]	valid_0's rmse: 1635.98
[2486]	valid_0's rmse: 1635.99
[2487]	valid_0's rmse: 1635.99
[2488]	valid_0's rmse: 1636.01
[2489]	valid_0's rmse: 1635.99
[2490]	valid_0's rmse: 1636.03
[2491]	valid_0's rmse: 1636.03
[2492]	valid_0's rmse: 1635.98
[2493]	valid_0's rmse: 1635.94
[2494]	valid_0's rmse: 1635.96
[2495]	valid_0's rmse: 1635.97
[2496]	valid_0's rmse: 1635.93
[2497]	valid_0's rmse: 1635.93
[2498]	valid_0's rmse: 1635.93
[2499]	valid_0's rmse: 1635.93
[2500]	valid_0's rmse: 1635.94
[2501]	valid_0's rmse: 1635.94
[2502]	valid_0's rmse: 1635.95
[2503]	valid_0's rmse: 1635.96
[2504]	valid_0's rmse: 1635.97
[2505]	valid_0's rmse: 1636.01
[2506]	valid_0's rmse: 1636.04
[2507]	valid_0's rmse: 1635.99
[2508]	valid_0's rmse: 1636.01
[2509]	valid_0's rmse: 1636.12
[2510]	valid_0's rmse: 1636.12
[2511]	valid_0's rmse: 1636.13
[2512]	valid_0's rmse: 1636.11
[2513]	valid_0's rmse: 1636.11
[2514]	valid_0's rmse: 1636.09
[2515]	valid_0's rmse: 1636.09
[2516]	valid_0's rmse: 1636.07
[2517]	valid_0's rmse: 1636.06
[2518]	valid_0's rmse: 1636.07
[2519]	valid_0's rmse: 1636.07
[2520]	valid_0's rmse: 1636.07
[2521]	valid_0's rmse: 1636.13
[2522]	valid_0's rmse: 1636.11
[2523]	valid_0's rmse: 1636.12
[2524]	valid_0's rmse: 1636.14
[2525]	valid_0's rmse: 1636.15
[2526]	valid_0's rmse: 1636.14
[2527]	valid_0's rmse: 1636.12
[2528]	valid_0's rmse: 1636.15
[2529]	valid_0's rmse: 1636.14
[2530]	valid_0's rmse: 1636.18
[2531]	valid_0's rmse: 1636.11
[2532]	valid_0's rmse: 1636.09
[2533]	valid_0's rmse: 1636.08
[2534]	valid_0's rmse: 1636.08
[2535]	valid_0's rmse: 1636.11
[2536]	valid_0's rmse: 1636.09
[2537]	valid_0's rmse: 1636.09
[2538]	valid_0's rmse: 1636.13
[2539]	valid_0's rmse: 1636.13
[2540]	valid_0's rmse: 1636.16
[2541]	valid_0's rmse: 1636.19
[2542]	valid_0's rmse: 1636.17
[2543]	valid_0's rmse: 1636.15
[2544]	valid_0's rmse: 1636.15
[2545]	valid_0's rmse: 1636.13
[2546]	valid_0's rmse: 1636.13
[2547]	valid_0's rmse: 1636.13
[2548]	valid_0's rmse: 1636.1
[2549]	valid_0's rmse: 1636.06
[2550]	valid_0's rmse: 1636.05
[2551]	valid_0's rmse: 1636.07
[2552]	valid_0's rmse: 1636.05
[2553]	valid_0's rmse: 1636.03
[2554]	valid_0's rmse: 1635.99
[2555]	valid_0's rmse: 1635.99
[2556]	valid_0's rmse: 1635.95
[2557]	valid_0's rmse: 1635.91
[2558]	valid_0's rmse: 1635.89
[2559]	valid_0's rmse: 1635.88
[2560]	valid_0's rmse: 1635.86
[2561]	valid_0's rmse: 1635.86
[2562]	valid_0's rmse: 1635.86
[2563]	valid_0's rmse: 1635.9
[2564]	valid_0's rmse: 1635.88
[2565]	valid_0's rmse: 1635.96
[2566]	valid_0's rmse: 1635.93
[2567]	valid_0's rmse: 1635.88
[2568]	valid_0's rmse: 1635.86
[2569]	valid_0's rmse: 1635.83
[2570]	valid_0's rmse: 1635.86
[2571]	valid_0's rmse: 1635.84
[2572]	valid_0's rmse: 1635.84
[2573]	valid_0's rmse: 1635.9
[2574]	valid_0's rmse: 1635.9
[2575]	valid_0's rmse: 1635.9
[2576]	valid_0's rmse: 1635.92
[2577]	valid_0's rmse: 1635.95
[2578]	valid_0's rmse: 1635.95
[2579]	valid_0's rmse: 1635.93
[2580]	valid_0's rmse: 1635.97
[2581]	valid_0's rmse: 1635.96
[2582]	valid_0's rmse: 1635.97
[2583]	valid_0's rmse: 1635.98
[2584]	valid_0's rmse: 1636.01
[2585]	valid_0's rmse: 1636.06
[2586]	valid_0's rmse: 1636
[2587]	valid_0's rmse: 1636.01
[2588]	valid_0's rmse: 1636.02
[2589]	valid_0's rmse: 1636.02
[2590]	valid_0's rmse: 1636.04
[2591]	valid_0's rmse: 1636.02
[2592]	valid_0's rmse: 1636.01
[2593]	valid_0's rmse: 1636.03
[2594]	valid_0's rmse: 1636
[2595]	valid_0's rmse: 1636
[2596]	valid_0's rmse: 1635.93
[2597]	valid_0's rmse: 1635.94
[2598]	valid_0's rmse: 1635.93
[2599]	valid_0's rmse: 1635.94
[2600]	valid_0's rmse: 1635.95
[2601]	valid_0's rmse: 1635.93
[2602]	valid_0's rmse: 1635.91
[2603]	valid_0's rmse: 1635.89
[2604]	valid_0's rmse: 1635.95
[2605]	valid_0's rmse: 1635.95
[2606]	valid_0's rmse: 1635.95
[2607]	valid_0's rmse: 1635.89
[2608]	valid_0's rmse: 1635.91
[2609]	valid_0's rmse: 1635.89
[2610]	valid_0's rmse: 1635.88
[2611]	valid_0's rmse: 1635.9
[2612]	valid_0's rmse: 1635.92
[2613]	valid_0's rmse: 1635.88
[2614]	valid_0's rmse: 1635.91
[2615]	valid_0's rmse: 1635.9
[2616]	valid_0's rmse: 1635.88
[2617]	valid_0's rmse: 1635.88
[2618]	valid_0's rmse: 1635.89
[2619]	valid_0's rmse: 1635.92
[2620]	valid_0's rmse: 1635.91
[2621]	valid_0's rmse: 1635.91
[2622]	valid_0's rmse: 1635.92
[2623]	valid_0's rmse: 1635.94
[2624]	valid_0's rmse: 1635.98
[2625]	valid_0's rmse: 1635.99
[2626]	valid_0's rmse: 1635.99
[2627]	valid_0's rmse: 1636.02
[2628]	valid_0's rmse: 1636.05
[2629]	valid_0's rmse: 1636.04
[2630]	valid_0's rmse: 1636.06
[2631]	valid_0's rmse: 1636.04
[2632]	valid_0's rmse: 1636
[2633]	valid_0's rmse: 1636.02
[2634]	valid_0's rmse: 1636.05
[2635]	valid_0's rmse: 1636.04
[2636]	valid_0's rmse: 1636
[2637]	valid_0's rmse: 1636.03
[2638]	valid_0's rmse: 1636.01
[2639]	valid_0's rmse: 1635.96
[2640]	valid_0's rmse: 1635.93
[2641]	valid_0's rmse: 1635.91
[2642]	valid_0's rmse: 1635.87
[2643]	valid_0's rmse: 1635.88
[2644]	valid_0's rmse: 1635.87
[2645]	valid_0's rmse: 1635.84
[2646]	valid_0's rmse: 1635.78
[2647]	valid_0's rmse: 1635.77
[2648]	valid_0's rmse: 1635.72
[2649]	valid_0's rmse: 1635.66
[2650]	valid_0's rmse: 1635.65
[2651]	valid_0's rmse: 1635.63
[2652]	valid_0's rmse: 1635.63
[2653]	valid_0's rmse: 1635.64
[2654]	valid_0's rmse: 1635.68
[2655]	valid_0's rmse: 1635.65
[2656]	valid_0's rmse: 1635.63
[2657]	valid_0's rmse: 1635.66
[2658]	valid_0's rmse: 1635.64
[2659]	valid_0's rmse: 1635.68
[2660]	valid_0's rmse: 1635.68
[2661]	valid_0's rmse: 1635.69
[2662]	valid_0's rmse: 1635.7
[2663]	valid_0's rmse: 1635.73
[2664]	valid_0's rmse: 1635.74
[2665]	valid_0's rmse: 1635.7
[2666]	valid_0's rmse: 1635.69
[2667]	valid_0's rmse: 1635.67
[2668]	valid_0's rmse: 1635.64
[2669]	valid_0's rmse: 1635.64
[2670]	valid_0's rmse: 1635.63
[2671]	valid_0's rmse: 1635.63
[2672]	valid_0's rmse: 1635.63
[2673]	valid_0's rmse: 1635.69
[2674]	valid_0's rmse: 1635.75
[2675]	valid_0's rmse: 1635.75
[2676]	valid_0's rmse: 1635.74
[2677]	valid_0's rmse: 1635.75
[2678]	valid_0's rmse: 1635.75
[2679]	valid_0's rmse: 1635.77
[2680]	valid_0's rmse: 1635.77
[2681]	valid_0's rmse: 1635.72
[2682]	valid_0's rmse: 1635.76
[2683]	valid_0's rmse: 1635.79
[2684]	valid_0's rmse: 1635.84
[2685]	valid_0's rmse: 1635.88
[2686]	valid_0's rmse: 1635.92
[2687]	valid_0's rmse: 1635.89
[2688]	valid_0's rmse: 1635.88
[2689]	valid_0's rmse: 1635.89
[2690]	valid_0's rmse: 1635.87
[2691]	valid_0's rmse: 1635.86
[2692]	valid_0's rmse: 1635.86
[2693]	valid_0's rmse: 1635.84
[2694]	valid_0's rmse: 1635.85
[2695]	valid_0's rmse: 1635.89
[2696]	valid_0's rmse: 1635.85
[2697]	valid_0's rmse: 1635.86
[2698]	valid_0's rmse: 1635.87
[2699]	valid_0's rmse: 1635.85
[2700]	valid_0's rmse: 1635.86
[2701]	valid_0's rmse: 1635.87
[2702]	valid_0's rmse: 1635.88
[2703]	valid_0's rmse: 1635.86
[2704]	valid_0's rmse: 1635.89
[2705]	valid_0's rmse: 1635.91
[2706]	valid_0's rmse: 1635.91
[2707]	valid_0's rmse: 1635.88
[2708]	valid_0's rmse: 1635.84
[2709]	valid_0's rmse: 1635.85
[2710]	valid_0's rmse: 1635.83
[2711]	valid_0's rmse: 1635.84
[2712]	valid_0's rmse: 1635.85
[2713]	valid_0's rmse: 1635.88
[2714]	valid_0's rmse: 1635.9
[2715]	valid_0's rmse: 1635.88
[2716]	valid_0's rmse: 1635.88
[2717]	valid_0's rmse: 1635.88
[2718]	valid_0's rmse: 1635.86
[2719]	valid_0's rmse: 1635.85
[2720]	valid_0's rmse: 1635.83
[2721]	valid_0's rmse: 1635.79
[2722]	valid_0's rmse: 1635.8
[2723]	valid_0's rmse: 1635.78
[2724]	valid_0's rmse: 1635.73
[2725]	valid_0's rmse: 1635.7
[2726]	valid_0's rmse: 1635.7
[2727]	valid_0's rmse: 1635.72
[2728]	valid_0's rmse: 1635.74
[2729]	valid_0's rmse: 1635.76
[2730]	valid_0's rmse: 1635.71
[2731]	valid_0's rmse: 1635.73
[2732]	valid_0's rmse: 1635.74
[2733]	valid_0's rmse: 1635.73
[2734]	valid_0's rmse: 1635.72
[2735]	valid_0's rmse: 1635.74
[2736]	valid_0's rmse: 1635.74
[2737]	valid_0's rmse: 1635.69
[2738]	valid_0's rmse: 1635.67
[2739]	valid_0's rmse: 1635.66
[2740]	valid_0's rmse: 1635.66
[2741]	valid_0's rmse: 1635.67
[2742]	valid_0's rmse: 1635.68
[2743]	valid_0's rmse: 1635.65
[2744]	valid_0's rmse: 1635.69
[2745]	valid_0's rmse: 1635.66
[2746]	valid_0's rmse: 1635.65
[2747]	valid_0's rmse: 1635.64
[2748]	valid_0's rmse: 1635.61
[2749]	valid_0's rmse: 1635.57
[2750]	valid_0's rmse: 1635.57
[2751]	valid_0's rmse: 1635.58
[2752]	valid_0's rmse: 1635.54
[2753]	valid_0's rmse: 1635.55
[2754]	valid_0's rmse: 1635.56
[2755]	valid_0's rmse: 1635.57
[2756]	valid_0's rmse: 1635.61
[2757]	valid_0's rmse: 1635.61
[2758]	valid_0's rmse: 1635.6
[2759]	valid_0's rmse: 1635.57
[2760]	valid_0's rmse: 1635.58
[2761]	valid_0's rmse: 1635.57
[2762]	valid_0's rmse: 1635.54
[2763]	valid_0's rmse: 1635.48
[2764]	valid_0's rmse: 1635.49
[2765]	valid_0's rmse: 1635.49
[2766]	valid_0's rmse: 1635.53
[2767]	valid_0's rmse: 1635.54
[2768]	valid_0's rmse: 1635.57
[2769]	valid_0's rmse: 1635.57
[2770]	valid_0's rmse: 1635.55
[2771]	valid_0's rmse: 1635.54
[2772]	valid_0's rmse: 1635.51
[2773]	valid_0's rmse: 1635.53
[2774]	valid_0's rmse: 1635.54
[2775]	valid_0's rmse: 1635.55
[2776]	valid_0's rmse: 1635.57
[2777]	valid_0's rmse: 1635.58
[2778]	valid_0's rmse: 1635.58
[2779]	valid_0's rmse: 1635.59
[2780]	valid_0's rmse: 1635.58
[2781]	valid_0's rmse: 1635.55
[2782]	valid_0's rmse: 1635.53
[2783]	valid_0's rmse: 1635.57
[2784]	valid_0's rmse: 1635.58
[2785]	valid_0's rmse: 1635.56
[2786]	valid_0's rmse: 1635.55
[2787]	valid_0's rmse: 1635.52
[2788]	valid_0's rmse: 1635.5
[2789]	valid_0's rmse: 1635.55
[2790]	valid_0's rmse: 1635.56
[2791]	valid_0's rmse: 1635.57
[2792]	valid_0's rmse: 1635.57
[2793]	valid_0's rmse: 1635.59
[2794]	valid_0's rmse: 1635.62
[2795]	valid_0's rmse: 1635.6
[2796]	valid_0's rmse: 1635.6
[2797]	valid_0's rmse: 1635.59
[2798]	valid_0's rmse: 1635.61
[2799]	valid_0's rmse: 1635.59
[2800]	valid_0's rmse: 1635.59
[2801]	valid_0's rmse: 1635.6
[2802]	valid_0's rmse: 1635.6
[2803]	valid_0's rmse: 1635.53
[2804]	valid_0's rmse: 1635.53
[2805]	valid_0's rmse: 1635.56
[2806]	valid_0's rmse: 1635.55
[2807]	valid_0's rmse: 1635.57
[2808]	valid_0's rmse: 1635.57
[2809]	valid_0's rmse: 1635.59
[2810]	valid_0's rmse: 1635.61
[2811]	valid_0's rmse: 1635.61
[2812]	valid_0's rmse: 1635.59
[2813]	valid_0's rmse: 1635.58
[2814]	valid_0's rmse: 1635.56
[2815]	valid_0's rmse: 1635.55
[2816]	valid_0's rmse: 1635.54
[2817]	valid_0's rmse: 1635.57
[2818]	valid_0's rmse: 1635.54
[2819]	valid_0's rmse: 1635.48
[2820]	valid_0's rmse: 1635.47
[2821]	valid_0's rmse: 1635.44
[2822]	valid_0's rmse: 1635.41
[2823]	valid_0's rmse: 1635.43
[2824]	valid_0's rmse: 1635.41
[2825]	valid_0's rmse: 1635.36
[2826]	valid_0's rmse: 1635.31
[2827]	valid_0's rmse: 1635.34
[2828]	valid_0's rmse: 1635.34
[2829]	valid_0's rmse: 1635.33
[2830]	valid_0's rmse: 1635.34
[2831]	valid_0's rmse: 1635.34
[2832]	valid_0's rmse: 1635.42
[2833]	valid_0's rmse: 1635.38
[2834]	valid_0's rmse: 1635.37
[2835]	valid_0's rmse: 1635.36
[2836]	valid_0's rmse: 1635.34
[2837]	valid_0's rmse: 1635.34
[2838]	valid_0's rmse: 1635.35
[2839]	valid_0's rmse: 1635.36
[2840]	valid_0's rmse: 1635.38
[2841]	valid_0's rmse: 1635.39
[2842]	valid_0's rmse: 1635.38
[2843]	valid_0's rmse: 1635.34
[2844]	valid_0's rmse: 1635.34
[2845]	valid_0's rmse: 1635.35
[2846]	valid_0's rmse: 1635.34
[2847]	valid_0's rmse: 1635.36
[2848]	valid_0's rmse: 1635.35
[2849]	valid_0's rmse: 1635.35
[2850]	valid_0's rmse: 1635.31
[2851]	valid_0's rmse: 1635.31
[2852]	valid_0's rmse: 1635.32
[2853]	valid_0's rmse: 1635.31
[2854]	valid_0's rmse: 1635.32
[2855]	valid_0's rmse: 1635.32
[2856]	valid_0's rmse: 1635.33
[2857]	valid_0's rmse: 1635.3
[2858]	valid_0's rmse: 1635.32
[2859]	valid_0's rmse: 1635.29
[2860]	valid_0's rmse: 1635.36
[2861]	valid_0's rmse: 1635.35
[2862]	valid_0's rmse: 1635.34
[2863]	valid_0's rmse: 1635.35
[2864]	valid_0's rmse: 1635.33
[2865]	valid_0's rmse: 1635.36
[2866]	valid_0's rmse: 1635.36
[2867]	valid_0's rmse: 1635.38
[2868]	valid_0's rmse: 1635.36
[2869]	valid_0's rmse: 1635.34
[2870]	valid_0's rmse: 1635.34
[2871]	valid_0's rmse: 1635.36
[2872]	valid_0's rmse: 1635.34
[2873]	valid_0's rmse: 1635.35
[2874]	valid_0's rmse: 1635.34
[2875]	valid_0's rmse: 1635.37
[2876]	valid_0's rmse: 1635.35
[2877]	valid_0's rmse: 1635.37
[2878]	valid_0's rmse: 1635.35
[2879]	valid_0's rmse: 1635.37
[2880]	valid_0's rmse: 1635.37
[2881]	valid_0's rmse: 1635.41
[2882]	valid_0's rmse: 1635.4
[2883]	valid_0's rmse: 1635.37
[2884]	valid_0's rmse: 1635.35
[2885]	valid_0's rmse: 1635.37
[2886]	valid_0's rmse: 1635.38
[2887]	valid_0's rmse: 1635.36
[2888]	valid_0's rmse: 1635.35
[2889]	valid_0's rmse: 1635.27
[2890]	valid_0's rmse: 1635.28
[2891]	valid_0's rmse: 1635.28
[2892]	valid_0's rmse: 1635.27
[2893]	valid_0's rmse: 1635.3
[2894]	valid_0's rmse: 1635.28
[2895]	valid_0's rmse: 1635.28
[2896]	valid_0's rmse: 1635.29
[2897]	valid_0's rmse: 1635.3
[2898]	valid_0's rmse: 1635.24
[2899]	valid_0's rmse: 1635.23
[2900]	valid_0's rmse: 1635.26
[2901]	valid_0's rmse: 1635.29
[2902]	valid_0's rmse: 1635.36
[2903]	valid_0's rmse: 1635.3
[2904]	valid_0's rmse: 1635.35
[2905]	valid_0's rmse: 1635.32
[2906]	valid_0's rmse: 1635.31
[2907]	valid_0's rmse: 1635.37
[2908]	valid_0's rmse: 1635.34
[2909]	valid_0's rmse: 1635.31
[2910]	valid_0's rmse: 1635.28
[2911]	valid_0's rmse: 1635.32
[2912]	valid_0's rmse: 1635.35
[2913]	valid_0's rmse: 1635.35
[2914]	valid_0's rmse: 1635.35
[2915]	valid_0's rmse: 1635.36
[2916]	valid_0's rmse: 1635.35
[2917]	valid_0's rmse: 1635.42
[2918]	valid_0's rmse: 1635.43
[2919]	valid_0's rmse: 1635.47
[2920]	valid_0's rmse: 1635.44
[2921]	valid_0's rmse: 1635.44
[2922]	valid_0's rmse: 1635.48
[2923]	valid_0's rmse: 1635.45
[2924]	valid_0's rmse: 1635.49
[2925]	valid_0's rmse: 1635.49
[2926]	valid_0's rmse: 1635.5
[2927]	valid_0's rmse: 1635.46
[2928]	valid_0's rmse: 1635.46
[2929]	valid_0's rmse: 1635.46
[2930]	valid_0's rmse: 1635.54
[2931]	valid_0's rmse: 1635.59
[2932]	valid_0's rmse: 1635.61
[2933]	valid_0's rmse: 1635.65
[2934]	valid_0's rmse: 1635.62
[2935]	valid_0's rmse: 1635.59
[2936]	valid_0's rmse: 1635.57
[2937]	valid_0's rmse: 1635.56
[2938]	valid_0's rmse: 1635.53
[2939]	valid_0's rmse: 1635.53
[2940]	valid_0's rmse: 1635.57
[2941]	valid_0's rmse: 1635.56
[2942]	valid_0's rmse: 1635.57
[2943]	valid_0's rmse: 1635.61
[2944]	valid_0's rmse: 1635.62
[2945]	valid_0's rmse: 1635.66
[2946]	valid_0's rmse: 1635.65
[2947]	valid_0's rmse: 1635.63
[2948]	valid_0's rmse: 1635.63
[2949]	valid_0's rmse: 1635.64
[2950]	valid_0's rmse: 1635.64
[2951]	valid_0's rmse: 1635.64
[2952]	valid_0's rmse: 1635.66
[2953]	valid_0's rmse: 1635.63
[2954]	valid_0's rmse: 1635.58
[2955]	valid_0's rmse: 1635.57
[2956]	valid_0's rmse: 1635.59
[2957]	valid_0's rmse: 1635.59
[2958]	valid_0's rmse: 1635.64
[2959]	valid_0's rmse: 1635.67
[2960]	valid_0's rmse: 1635.67
[2961]	valid_0's rmse: 1635.66
[2962]	valid_0's rmse: 1635.66
[2963]	valid_0's rmse: 1635.64
[2964]	valid_0's rmse: 1635.62
[2965]	valid_0's rmse: 1635.62
[2966]	valid_0's rmse: 1635.62
[2967]	valid_0's rmse: 1635.63
[2968]	valid_0's rmse: 1635.61
[2969]	valid_0's rmse: 1635.6
[2970]	valid_0's rmse: 1635.58
[2971]	valid_0's rmse: 1635.58
[2972]	valid_0's rmse: 1635.51
[2973]	valid_0's rmse: 1635.51
[2974]	valid_0's rmse: 1635.51
[2975]	valid_0's rmse: 1635.5
[2976]	valid_0's rmse: 1635.48
[2977]	valid_0's rmse: 1635.46
[2978]	valid_0's rmse: 1635.44
[2979]	valid_0's rmse: 1635.41
[2980]	valid_0's rmse: 1635.44
[2981]	valid_0's rmse: 1635.44
[2982]	valid_0's rmse: 1635.39
[2983]	valid_0's rmse: 1635.4
[2984]	valid_0's rmse: 1635.42
[2985]	valid_0's rmse: 1635.4
[2986]	valid_0's rmse: 1635.39
[2987]	valid_0's rmse: 1635.38
[2988]	valid_0's rmse: 1635.34
[2989]	valid_0's rmse: 1635.32
[2990]	valid_0's rmse: 1635.36
[2991]	valid_0's rmse: 1635.36
[2992]	valid_0's rmse: 1635.35
[2993]	valid_0's rmse: 1635.33
[2994]	valid_0's rmse: 1635.28
[2995]	valid_0's rmse: 1635.3
[2996]	valid_0's rmse: 1635.3
[2997]	valid_0's rmse: 1635.28
[2998]	valid_0's rmse: 1635.27
[2999]	valid_0's rmse: 1635.26
[3000]	valid_0's rmse: 1635.26
[3001]	valid_0's rmse: 1635.26
[3002]	valid_0's rmse: 1635.24
[3003]	valid_0's rmse: 1635.22
[3004]	valid_0's rmse: 1635.18
[3005]	valid_0's rmse: 1635.17
[3006]	valid_0's rmse: 1635.15
[3007]	valid_0's rmse: 1635.15
[3008]	valid_0's rmse: 1635.1
[3009]	valid_0's rmse: 1635.1
[3010]	valid_0's rmse: 1635.11
[3011]	valid_0's rmse: 1635.12
[3012]	valid_0's rmse: 1635.13
[3013]	valid_0's rmse: 1635.12
[3014]	valid_0's rmse: 1635.14
[3015]	valid_0's rmse: 1635.13
[3016]	valid_0's rmse: 1635.13
[3017]	valid_0's rmse: 1635.14
[3018]	valid_0's rmse: 1635.13
[3019]	valid_0's rmse: 1635.12
[3020]	valid_0's rmse: 1635.1
[3021]	valid_0's rmse: 1635.1
[3022]	valid_0's rmse: 1635.1
[3023]	valid_0's rmse: 1635.1
[3024]	valid_0's rmse: 1635.09
[3025]	valid_0's rmse: 1635.12
[3026]	valid_0's rmse: 1635.13
[3027]	valid_0's rmse: 1635.13
[3028]	valid_0's rmse: 1635.15
[3029]	valid_0's rmse: 1635.14
[3030]	valid_0's rmse: 1635.13
[3031]	valid_0's rmse: 1635.17
[3032]	valid_0's rmse: 1635.18
[3033]	valid_0's rmse: 1635.17
[3034]	valid_0's rmse: 1635.17
[3035]	valid_0's rmse: 1635.19
[3036]	valid_0's rmse: 1635.23
[3037]	valid_0's rmse: 1635.21
[3038]	valid_0's rmse: 1635.25
[3039]	valid_0's rmse: 1635.25
[3040]	valid_0's rmse: 1635.19
[3041]	valid_0's rmse: 1635.27
[3042]	valid_0's rmse: 1635.27
[3043]	valid_0's rmse: 1635.31
[3044]	valid_0's rmse: 1635.33
[3045]	valid_0's rmse: 1635.33
[3046]	valid_0's rmse: 1635.32
[3047]	valid_0's rmse: 1635.3
[3048]	valid_0's rmse: 1635.33
[3049]	valid_0's rmse: 1635.36
[3050]	valid_0's rmse: 1635.4
[3051]	valid_0's rmse: 1635.4
[3052]	valid_0's rmse: 1635.42
[3053]	valid_0's rmse: 1635.39
[3054]	valid_0's rmse: 1635.35
[3055]	valid_0's rmse: 1635.38
[3056]	valid_0's rmse: 1635.37
[3057]	valid_0's rmse: 1635.36
[3058]	valid_0's rmse: 1635.37
[3059]	valid_0's rmse: 1635.36
[3060]	valid_0's rmse: 1635.37
[3061]	valid_0's rmse: 1635.37
[3062]	valid_0's rmse: 1635.32
[3063]	valid_0's rmse: 1635.31
[3064]	valid_0's rmse: 1635.35
[3065]	valid_0's rmse: 1635.35
[3066]	valid_0's rmse: 1635.37
[3067]	valid_0's rmse: 1635.38
[3068]	valid_0's rmse: 1635.41
[3069]	valid_0's rmse: 1635.43
[3070]	valid_0's rmse: 1635.41
[3071]	valid_0's rmse: 1635.45
[3072]	valid_0's rmse: 1635.46
[3073]	valid_0's rmse: 1635.47
[3074]	valid_0's rmse: 1635.46
[3075]	valid_0's rmse: 1635.46
[3076]	valid_0's rmse: 1635.47
[3077]	valid_0's rmse: 1635.46
[3078]	valid_0's rmse: 1635.43
[3079]	valid_0's rmse: 1635.45
[3080]	valid_0's rmse: 1635.46
[3081]	valid_0's rmse: 1635.48
[3082]	valid_0's rmse: 1635.47
[3083]	valid_0's rmse: 1635.49
[3084]	valid_0's rmse: 1635.46
[3085]	valid_0's rmse: 1635.48
[3086]	valid_0's rmse: 1635.51
[3087]	valid_0's rmse: 1635.51
[3088]	valid_0's rmse: 1635.48
[3089]	valid_0's rmse: 1635.5
[3090]	valid_0's rmse: 1635.51
[3091]	valid_0's rmse: 1635.52
[3092]	valid_0's rmse: 1635.54
[3093]	valid_0's rmse: 1635.53
[3094]	valid_0's rmse: 1635.5
[3095]	valid_0's rmse: 1635.5
[3096]	valid_0's rmse: 1635.5
[3097]	valid_0's rmse: 1635.5
[3098]	valid_0's rmse: 1635.51
[3099]	valid_0's rmse: 1635.5
[3100]	valid_0's rmse: 1635.54
[3101]	valid_0's rmse: 1635.54
[3102]	valid_0's rmse: 1635.54
[3103]	valid_0's rmse: 1635.56
[3104]	valid_0's rmse: 1635.56
[3105]	valid_0's rmse: 1635.5
[3106]	valid_0's rmse: 1635.46
[3107]	valid_0's rmse: 1635.44
[3108]	valid_0's rmse: 1635.4
[3109]	valid_0's rmse: 1635.39
[3110]	valid_0's rmse: 1635.39
[3111]	valid_0's rmse: 1635.4
[3112]	valid_0's rmse: 1635.4
[3113]	valid_0's rmse: 1635.35
[3114]	valid_0's rmse: 1635.33
[3115]	valid_0's rmse: 1635.27
[3116]	valid_0's rmse: 1635.25
[3117]	valid_0's rmse: 1635.22
[3118]	valid_0's rmse: 1635.21
[3119]	valid_0's rmse: 1635.21
[3120]	valid_0's rmse: 1635.22
[3121]	valid_0's rmse: 1635.21
[3122]	valid_0's rmse: 1635.18
[3123]	valid_0's rmse: 1635.15
[3124]	valid_0's rmse: 1635.15
[3125]	valid_0's rmse: 1635.14
[3126]	valid_0's rmse: 1635.12
[3127]	valid_0's rmse: 1635.1
[3128]	valid_0's rmse: 1635.15
[3129]	valid_0's rmse: 1635.14
[3130]	valid_0's rmse: 1635.1
[3131]	valid_0's rmse: 1635.08
[3132]	valid_0's rmse: 1635.07
[3133]	valid_0's rmse: 1635.1
[3134]	valid_0's rmse: 1635.09
[3135]	valid_0's rmse: 1635.07
[3136]	valid_0's rmse: 1635.05
[3137]	valid_0's rmse: 1635.05
[3138]	valid_0's rmse: 1635.09
[3139]	valid_0's rmse: 1635.08
[3140]	valid_0's rmse: 1635.1
[3141]	valid_0's rmse: 1635.08
[3142]	valid_0's rmse: 1635.09
[3143]	valid_0's rmse: 1635.11
[3144]	valid_0's rmse: 1635.11
[3145]	valid_0's rmse: 1635.11
[3146]	valid_0's rmse: 1635.12
[3147]	valid_0's rmse: 1635.12
[3148]	valid_0's rmse: 1635.13
[3149]	valid_0's rmse: 1635.13
[3150]	valid_0's rmse: 1635.17
[3151]	valid_0's rmse: 1635.16
[3152]	valid_0's rmse: 1635.2
[3153]	valid_0's rmse: 1635.22
[3154]	valid_0's rmse: 1635.23
[3155]	valid_0's rmse: 1635.19
[3156]	valid_0's rmse: 1635.16
[3157]	valid_0's rmse: 1635.15
[3158]	valid_0's rmse: 1635.16
[3159]	valid_0's rmse: 1635.15
[3160]	valid_0's rmse: 1635.14
[3161]	valid_0's rmse: 1635.15
[3162]	valid_0's rmse: 1635.09
[3163]	valid_0's rmse: 1635.09
[3164]	valid_0's rmse: 1635.07
[3165]	valid_0's rmse: 1635.07
[3166]	valid_0's rmse: 1635.07
[3167]	valid_0's rmse: 1635.06
[3168]	valid_0's rmse: 1635.07
[3169]	valid_0's rmse: 1635.07
[3170]	valid_0's rmse: 1635.07
[3171]	valid_0's rmse: 1635.11
[3172]	valid_0's rmse: 1635.13
[3173]	valid_0's rmse: 1635.13
[3174]	valid_0's rmse: 1635.1
[3175]	valid_0's rmse: 1635.1
[3176]	valid_0's rmse: 1635.1
[3177]	valid_0's rmse: 1635.11
[3178]	valid_0's rmse: 1635.16
[3179]	valid_0's rmse: 1635.15
[3180]	valid_0's rmse: 1635.16
[3181]	valid_0's rmse: 1635.14
[3182]	valid_0's rmse: 1635.14
[3183]	valid_0's rmse: 1635.18
[3184]	valid_0's rmse: 1635.21
[3185]	valid_0's rmse: 1635.21
[3186]	valid_0's rmse: 1635.17
[3187]	valid_0's rmse: 1635.15
[3188]	valid_0's rmse: 1635.17
[3189]	valid_0's rmse: 1635.2
[3190]	valid_0's rmse: 1635.21
[3191]	valid_0's rmse: 1635.19
[3192]	valid_0's rmse: 1635.2
[3193]	valid_0's rmse: 1635.2
[3194]	valid_0's rmse: 1635.21
[3195]	valid_0's rmse: 1635.2
[3196]	valid_0's rmse: 1635.21
[3197]	valid_0's rmse: 1635.2
[3198]	valid_0's rmse: 1635.22
[3199]	valid_0's rmse: 1635.19
[3200]	valid_0's rmse: 1635.21
[3201]	valid_0's rmse: 1635.21
[3202]	valid_0's rmse: 1635.2
[3203]	valid_0's rmse: 1635.16
[3204]	valid_0's rmse: 1635.16
[3205]	valid_0's rmse: 1635.15
[3206]	valid_0's rmse: 1635.11
[3207]	valid_0's rmse: 1635.04
[3208]	valid_0's rmse: 1635.03
[3209]	valid_0's rmse: 1635.02
[3210]	valid_0's rmse: 1635.02
[3211]	valid_0's rmse: 1635.02
[3212]	valid_0's rmse: 1634.98
[3213]	valid_0's rmse: 1634.99
[3214]	valid_0's rmse: 1635.01
[3215]	valid_0's rmse: 1635.04
[3216]	valid_0's rmse: 1635.09
[3217]	valid_0's rmse: 1635.07
[3218]	valid_0's rmse: 1635.09
[3219]	valid_0's rmse: 1635.09
[3220]	valid_0's rmse: 1635.09
[3221]	valid_0's rmse: 1635.12
[3222]	valid_0's rmse: 1635.19
[3223]	valid_0's rmse: 1635.19
[3224]	valid_0's rmse: 1635.19
[3225]	valid_0's rmse: 1635.22
[3226]	valid_0's rmse: 1635.21
[3227]	valid_0's rmse: 1635.24
[3228]	valid_0's rmse: 1635.27
[3229]	valid_0's rmse: 1635.32
[3230]	valid_0's rmse: 1635.32
[3231]	valid_0's rmse: 1635.31
[3232]	valid_0's rmse: 1635.33
[3233]	valid_0's rmse: 1635.32
[3234]	valid_0's rmse: 1635.33
[3235]	valid_0's rmse: 1635.33
[3236]	valid_0's rmse: 1635.34
[3237]	valid_0's rmse: 1635.37
[3238]	valid_0's rmse: 1635.39
[3239]	valid_0's rmse: 1635.4
[3240]	valid_0's rmse: 1635.4
[3241]	valid_0's rmse: 1635.41
[3242]	valid_0's rmse: 1635.42
[3243]	valid_0's rmse: 1635.4
[3244]	valid_0's rmse: 1635.41
[3245]	valid_0's rmse: 1635.46
[3246]	valid_0's rmse: 1635.48
[3247]	valid_0's rmse: 1635.47
[3248]	valid_0's rmse: 1635.47
[3249]	valid_0's rmse: 1635.46
[3250]	valid_0's rmse: 1635.48
[3251]	valid_0's rmse: 1635.48
[3252]	valid_0's rmse: 1635.48
[3253]	valid_0's rmse: 1635.51
[3254]	valid_0's rmse: 1635.52
[3255]	valid_0's rmse: 1635.5
[3256]	valid_0's rmse: 1635.51
[3257]	valid_0's rmse: 1635.5
[3258]	valid_0's rmse: 1635.5
[3259]	valid_0's rmse: 1635.46
[3260]	valid_0's rmse: 1635.45
[3261]	valid_0's rmse: 1635.45
[3262]	valid_0's rmse: 1635.42
[3263]	valid_0's rmse: 1635.41
[3264]	valid_0's rmse: 1635.39
[3265]	valid_0's rmse: 1635.38
[3266]	valid_0's rmse: 1635.4
[3267]	valid_0's rmse: 1635.39
[3268]	valid_0's rmse: 1635.38
[3269]	valid_0's rmse: 1635.34
[3270]	valid_0's rmse: 1635.34
[3271]	valid_0's rmse: 1635.35
[3272]	valid_0's rmse: 1635.34
[3273]	valid_0's rmse: 1635.31
[3274]	valid_0's rmse: 1635.3
[3275]	valid_0's rmse: 1635.3
[3276]	valid_0's rmse: 1635.29
[3277]	valid_0's rmse: 1635.26
[3278]	valid_0's rmse: 1635.23
[3279]	valid_0's rmse: 1635.26
[3280]	valid_0's rmse: 1635.26
[3281]	valid_0's rmse: 1635.25
[3282]	valid_0's rmse: 1635.21
[3283]	valid_0's rmse: 1635.25
[3284]	valid_0's rmse: 1635.25
[3285]	valid_0's rmse: 1635.28
[3286]	valid_0's rmse: 1635.27
[3287]	valid_0's rmse: 1635.24
[3288]	valid_0's rmse: 1635.21
[3289]	valid_0's rmse: 1635.19
[3290]	valid_0's rmse: 1635.13
[3291]	valid_0's rmse: 1635.11
[3292]	valid_0's rmse: 1635.14
[3293]	valid_0's rmse: 1635.11
[3294]	valid_0's rmse: 1635.11
[3295]	valid_0's rmse: 1635.12
[3296]	valid_0's rmse: 1635.12
[3297]	valid_0's rmse: 1635.08
[3298]	valid_0's rmse: 1635.08
[3299]	valid_0's rmse: 1635.06
[3300]	valid_0's rmse: 1635.06
[3301]	valid_0's rmse: 1635.07
[3302]	valid_0's rmse: 1635.02
[3303]	valid_0's rmse: 1635.01
[3304]	valid_0's rmse: 1635
[3305]	valid_0's rmse: 1634.99
[3306]	valid_0's rmse: 1634.98
[3307]	valid_0's rmse: 1635.03
[3308]	valid_0's rmse: 1635
[3309]	valid_0's rmse: 1634.98
[3310]	valid_0's rmse: 1634.99
[3311]	valid_0's rmse: 1634.99
[3312]	valid_0's rmse: 1635
[3313]	valid_0's rmse: 1634.97
[3314]	valid_0's rmse: 1635.02
[3315]	valid_0's rmse: 1635.01
[3316]	valid_0's rmse: 1634.99
[3317]	valid_0's rmse: 1634.99
[3318]	valid_0's rmse: 1634.97
[3319]	valid_0's rmse: 1634.97
[3320]	valid_0's rmse: 1635
[3321]	valid_0's rmse: 1635.02
[3322]	valid_0's rmse: 1635.06
[3323]	valid_0's rmse: 1635.08
[3324]	valid_0's rmse: 1635.1
[3325]	valid_0's rmse: 1635.08
[3326]	valid_0's rmse: 1635.11
[3327]	valid_0's rmse: 1635.09
[3328]	valid_0's rmse: 1635.14
[3329]	valid_0's rmse: 1635.15
[3330]	valid_0's rmse: 1635.12
[3331]	valid_0's rmse: 1635.11
[3332]	valid_0's rmse: 1635.1
[3333]	valid_0's rmse: 1635.12
[3334]	valid_0's rmse: 1635.11
[3335]	valid_0's rmse: 1635.1
[3336]	valid_0's rmse: 1635.09
[3337]	valid_0's rmse: 1635.08
[3338]	valid_0's rmse: 1635.08
[3339]	valid_0's rmse: 1635.1
[3340]	valid_0's rmse: 1635.07
[3341]	valid_0's rmse: 1635.06
[3342]	valid_0's rmse: 1635.04
[3343]	valid_0's rmse: 1635.02
[3344]	valid_0's rmse: 1635.01
[3345]	valid_0's rmse: 1635.02
[3346]	valid_0's rmse: 1635.04
[3347]	valid_0's rmse: 1635.02
[3348]	valid_0's rmse: 1634.97
[3349]	valid_0's rmse: 1634.95
[3350]	valid_0's rmse: 1634.97
[3351]	valid_0's rmse: 1634.95
[3352]	valid_0's rmse: 1634.93
[3353]	valid_0's rmse: 1634.91
[3354]	valid_0's rmse: 1634.88
[3355]	valid_0's rmse: 1634.89
[3356]	valid_0's rmse: 1634.83
[3357]	valid_0's rmse: 1634.84
[3358]	valid_0's rmse: 1634.84
[3359]	valid_0's rmse: 1634.84
[3360]	valid_0's rmse: 1634.87
[3361]	valid_0's rmse: 1634.88
[3362]	valid_0's rmse: 1634.82
[3363]	valid_0's rmse: 1634.82
[3364]	valid_0's rmse: 1634.78
[3365]	valid_0's rmse: 1634.83
[3366]	valid_0's rmse: 1634.86
[3367]	valid_0's rmse: 1634.87
[3368]	valid_0's rmse: 1634.87
[3369]	valid_0's rmse: 1634.88
[3370]	valid_0's rmse: 1634.88
[3371]	valid_0's rmse: 1634.88
[3372]	valid_0's rmse: 1634.9
[3373]	valid_0's rmse: 1634.92
[3374]	valid_0's rmse: 1634.89
[3375]	valid_0's rmse: 1634.91
[3376]	valid_0's rmse: 1634.88
[3377]	valid_0's rmse: 1634.92
[3378]	valid_0's rmse: 1634.9
[3379]	valid_0's rmse: 1634.89
[3380]	valid_0's rmse: 1634.88
[3381]	valid_0's rmse: 1634.9
[3382]	valid_0's rmse: 1634.94
[3383]	valid_0's rmse: 1634.93
[3384]	valid_0's rmse: 1634.9
[3385]	valid_0's rmse: 1634.9
[3386]	valid_0's rmse: 1634.9
[3387]	valid_0's rmse: 1634.9
[3388]	valid_0's rmse: 1634.89
[3389]	valid_0's rmse: 1634.88
[3390]	valid_0's rmse: 1634.9
[3391]	valid_0's rmse: 1634.9
[3392]	valid_0's rmse: 1634.86
[3393]	valid_0's rmse: 1634.86
[3394]	valid_0's rmse: 1634.85
[3395]	valid_0's rmse: 1634.83
[3396]	valid_0's rmse: 1634.83
[3397]	valid_0's rmse: 1634.89
[3398]	valid_0's rmse: 1634.89
[3399]	valid_0's rmse: 1634.9
[3400]	valid_0's rmse: 1634.89
[3401]	valid_0's rmse: 1634.88
[3402]	valid_0's rmse: 1634.86
[3403]	valid_0's rmse: 1634.82
[3404]	valid_0's rmse: 1634.81
[3405]	valid_0's rmse: 1634.84
[3406]	valid_0's rmse: 1634.79
[3407]	valid_0's rmse: 1634.81
[3408]	valid_0's rmse: 1634.81
[3409]	valid_0's rmse: 1634.8
[3410]	valid_0's rmse: 1634.83
[3411]	valid_0's rmse: 1634.84
[3412]	valid_0's rmse: 1634.85
[3413]	valid_0's rmse: 1634.9
[3414]	valid_0's rmse: 1634.9
[3415]	valid_0's rmse: 1634.91
[3416]	valid_0's rmse: 1634.93
[3417]	valid_0's rmse: 1634.9
[3418]	valid_0's rmse: 1634.89
[3419]	valid_0's rmse: 1634.87
[3420]	valid_0's rmse: 1634.85
[3421]	valid_0's rmse: 1634.85
[3422]	valid_0's rmse: 1634.89
[3423]	valid_0's rmse: 1634.87
[3424]	valid_0's rmse: 1634.86
[3425]	valid_0's rmse: 1634.88
[3426]	valid_0's rmse: 1634.85
[3427]	valid_0's rmse: 1634.85
[3428]	valid_0's rmse: 1634.85
[3429]	valid_0's rmse: 1634.85
[3430]	valid_0's rmse: 1634.86
[3431]	valid_0's rmse: 1634.84
[3432]	valid_0's rmse: 1634.86
[3433]	valid_0's rmse: 1634.85
[3434]	valid_0's rmse: 1634.84
[3435]	valid_0's rmse: 1634.85
[3436]	valid_0's rmse: 1634.85
[3437]	valid_0's rmse: 1634.84
[3438]	valid_0's rmse: 1634.88
[3439]	valid_0's rmse: 1634.91
[3440]	valid_0's rmse: 1634.91
[3441]	valid_0's rmse: 1634.97
[3442]	valid_0's rmse: 1634.92
[3443]	valid_0's rmse: 1634.94
[3444]	valid_0's rmse: 1634.91
[3445]	valid_0's rmse: 1634.89
[3446]	valid_0's rmse: 1634.89
[3447]	valid_0's rmse: 1634.88
[3448]	valid_0's rmse: 1634.87
[3449]	valid_0's rmse: 1634.88
[3450]	valid_0's rmse: 1634.9
[3451]	valid_0's rmse: 1634.91
[3452]	valid_0's rmse: 1634.94
[3453]	valid_0's rmse: 1634.96
[3454]	valid_0's rmse: 1634.91
[3455]	valid_0's rmse: 1634.91
[3456]	valid_0's rmse: 1634.94
[3457]	valid_0's rmse: 1634.96
[3458]	valid_0's rmse: 1634.94
[3459]	valid_0's rmse: 1634.96
[3460]	valid_0's rmse: 1634.96
[3461]	valid_0's rmse: 1634.98
[3462]	valid_0's rmse: 1634.99
[3463]	valid_0's rmse: 1634.99
[3464]	valid_0's rmse: 1634.98
[3465]	valid_0's rmse: 1634.97
[3466]	valid_0's rmse: 1635.01
[3467]	valid_0's rmse: 1635
[3468]	valid_0's rmse: 1634.97
[3469]	valid_0's rmse: 1634.99
[3470]	valid_0's rmse: 1635
[3471]	valid_0's rmse: 1635
[3472]	valid_0's rmse: 1635
[3473]	valid_0's rmse: 1635
[3474]	valid_0's rmse: 1634.96
[3475]	valid_0's rmse: 1634.94
[3476]	valid_0's rmse: 1634.94
[3477]	valid_0's rmse: 1634.95
[3478]	valid_0's rmse: 1634.94
[3479]	valid_0's rmse: 1634.94
[3480]	valid_0's rmse: 1634.95
[3481]	valid_0's rmse: 1634.93
[3482]	valid_0's rmse: 1634.86
[3483]	valid_0's rmse: 1634.87
[3484]	valid_0's rmse: 1634.9
[3485]	valid_0's rmse: 1634.9
[3486]	valid_0's rmse: 1634.86
[3487]	valid_0's rmse: 1634.85
[3488]	valid_0's rmse: 1634.83
[3489]	valid_0's rmse: 1634.87
[3490]	valid_0's rmse: 1634.89
[3491]	valid_0's rmse: 1634.91
[3492]	valid_0's rmse: 1634.92
[3493]	valid_0's rmse: 1634.96
[3494]	valid_0's rmse: 1634.97
[3495]	valid_0's rmse: 1634.96
[3496]	valid_0's rmse: 1634.96
[3497]	valid_0's rmse: 1634.95
[3498]	valid_0's rmse: 1634.94
[3499]	valid_0's rmse: 1634.95
[3500]	valid_0's rmse: 1634.94
[3501]	valid_0's rmse: 1634.95
[3502]	valid_0's rmse: 1634.95
[3503]	valid_0's rmse: 1634.9
[3504]	valid_0's rmse: 1634.89
[3505]	valid_0's rmse: 1634.88
[3506]	valid_0's rmse: 1634.87
[3507]	valid_0's rmse: 1634.87
[3508]	valid_0's rmse: 1634.87
[3509]	valid_0's rmse: 1634.86
[3510]	valid_0's rmse: 1634.87
[3511]	valid_0's rmse: 1634.86
[3512]	valid_0's rmse: 1634.88
[3513]	valid_0's rmse: 1634.9
[3514]	valid_0's rmse: 1634.93
[3515]	valid_0's rmse: 1634.96
[3516]	valid_0's rmse: 1634.96
[3517]	valid_0's rmse: 1634.97
[3518]	valid_0's rmse: 1634.98
[3519]	valid_0's rmse: 1634.96
[3520]	valid_0's rmse: 1634.93
[3521]	valid_0's rmse: 1634.93
[3522]	valid_0's rmse: 1634.93
[3523]	valid_0's rmse: 1634.93
[3524]	valid_0's rmse: 1634.92
[3525]	valid_0's rmse: 1634.89
[3526]	valid_0's rmse: 1634.89
[3527]	valid_0's rmse: 1634.89
[3528]	valid_0's rmse: 1634.89
[3529]	valid_0's rmse: 1634.88
[3530]	valid_0's rmse: 1634.91
[3531]	valid_0's rmse: 1634.94
[3532]	valid_0's rmse: 1634.92
[3533]	valid_0's rmse: 1634.94
[3534]	valid_0's rmse: 1634.94
[3535]	valid_0's rmse: 1634.96
[3536]	valid_0's rmse: 1634.99
[3537]	valid_0's rmse: 1635.05
[3538]	valid_0's rmse: 1635.06
[3539]	valid_0's rmse: 1635.06
[3540]	valid_0's rmse: 1635.06
[3541]	valid_0's rmse: 1635.07
[3542]	valid_0's rmse: 1635.08
[3543]	valid_0's rmse: 1635.07
[3544]	valid_0's rmse: 1635.1
[3545]	valid_0's rmse: 1635.08
[3546]	valid_0's rmse: 1635.09
[3547]	valid_0's rmse: 1635.1
[3548]	valid_0's rmse: 1635.1
[3549]	valid_0's rmse: 1635.07
[3550]	valid_0's rmse: 1635.06
[3551]	valid_0's rmse: 1635.07
[3552]	valid_0's rmse: 1635.07
[3553]	valid_0's rmse: 1635.06
[3554]	valid_0's rmse: 1635.05
[3555]	valid_0's rmse: 1635.06
[3556]	valid_0's rmse: 1635
[3557]	valid_0's rmse: 1635
[3558]	valid_0's rmse: 1634.97
[3559]	valid_0's rmse: 1634.99
[3560]	valid_0's rmse: 1634.99
[3561]	valid_0's rmse: 1634.98
[3562]	valid_0's rmse: 1634.97
[3563]	valid_0's rmse: 1634.98
[3564]	valid_0's rmse: 1634.99
[3565]	valid_0's rmse: 1634.96
[3566]	valid_0's rmse: 1634.94
[3567]	valid_0's rmse: 1634.96
[3568]	valid_0's rmse: 1635
[3569]	valid_0's rmse: 1634.99
[3570]	valid_0's rmse: 1635
[3571]	valid_0's rmse: 1635.02
[3572]	valid_0's rmse: 1635.04
[3573]	valid_0's rmse: 1635.03
[3574]	valid_0's rmse: 1635.03
[3575]	valid_0's rmse: 1635.04
[3576]	valid_0's rmse: 1635
[3577]	valid_0's rmse: 1634.97
[3578]	valid_0's rmse: 1635
[3579]	valid_0's rmse: 1634.99
[3580]	valid_0's rmse: 1634.97
[3581]	valid_0's rmse: 1634.98
[3582]	valid_0's rmse: 1634.98
[3583]	valid_0's rmse: 1635.01
[3584]	valid_0's rmse: 1635.04
[3585]	valid_0's rmse: 1635.01
[3586]	valid_0's rmse: 1634.99
[3587]	valid_0's rmse: 1634.99
[3588]	valid_0's rmse: 1634.96
[3589]	valid_0's rmse: 1634.89
[3590]	valid_0's rmse: 1634.88
[3591]	valid_0's rmse: 1634.87
[3592]	valid_0's rmse: 1634.87
[3593]	valid_0's rmse: 1634.85
[3594]	valid_0's rmse: 1634.83
[3595]	valid_0's rmse: 1634.86
[3596]	valid_0's rmse: 1634.88
[3597]	valid_0's rmse: 1634.88
[3598]	valid_0's rmse: 1634.83
[3599]	valid_0's rmse: 1634.82
[3600]	valid_0's rmse: 1634.82
[3601]	valid_0's rmse: 1634.84
[3602]	valid_0's rmse: 1634.83
[3603]	valid_0's rmse: 1634.81
[3604]	valid_0's rmse: 1634.82
[3605]	valid_0's rmse: 1634.78
[3606]	valid_0's rmse: 1634.79
[3607]	valid_0's rmse: 1634.78
[3608]	valid_0's rmse: 1634.74
[3609]	valid_0's rmse: 1634.73
[3610]	valid_0's rmse: 1634.69
[3611]	valid_0's rmse: 1634.69
[3612]	valid_0's rmse: 1634.7
[3613]	valid_0's rmse: 1634.69
[3614]	valid_0's rmse: 1634.7
[3615]	valid_0's rmse: 1634.74
[3616]	valid_0's rmse: 1634.74
[3617]	valid_0's rmse: 1634.74
[3618]	valid_0's rmse: 1634.77
[3619]	valid_0's rmse: 1634.83
[3620]	valid_0's rmse: 1634.82
[3621]	valid_0's rmse: 1634.84
[3622]	valid_0's rmse: 1634.8
[3623]	valid_0's rmse: 1634.81
[3624]	valid_0's rmse: 1634.85
[3625]	valid_0's rmse: 1634.82
[3626]	valid_0's rmse: 1634.82
[3627]	valid_0's rmse: 1634.81
[3628]	valid_0's rmse: 1634.84
[3629]	valid_0's rmse: 1634.83
[3630]	valid_0's rmse: 1634.79
[3631]	valid_0's rmse: 1634.79
[3632]	valid_0's rmse: 1634.8
[3633]	valid_0's rmse: 1634.82
[3634]	valid_0's rmse: 1634.78
[3635]	valid_0's rmse: 1634.81
[3636]	valid_0's rmse: 1634.81
[3637]	valid_0's rmse: 1634.8
[3638]	valid_0's rmse: 1634.84
[3639]	valid_0's rmse: 1634.84
[3640]	valid_0's rmse: 1634.77
[3641]	valid_0's rmse: 1634.79
[3642]	valid_0's rmse: 1634.77
[3643]	valid_0's rmse: 1634.78
[3644]	valid_0's rmse: 1634.77
[3645]	valid_0's rmse: 1634.74
[3646]	valid_0's rmse: 1634.74
[3647]	valid_0's rmse: 1634.74
[3648]	valid_0's rmse: 1634.76
[3649]	valid_0's rmse: 1634.72
[3650]	valid_0's rmse: 1634.71
[3651]	valid_0's rmse: 1634.64
[3652]	valid_0's rmse: 1634.63
[3653]	valid_0's rmse: 1634.64
[3654]	valid_0's rmse: 1634.63
[3655]	valid_0's rmse: 1634.62
[3656]	valid_0's rmse: 1634.63
[3657]	valid_0's rmse: 1634.61
[3658]	valid_0's rmse: 1634.6
[3659]	valid_0's rmse: 1634.64
[3660]	valid_0's rmse: 1634.65
[3661]	valid_0's rmse: 1634.65
[3662]	valid_0's rmse: 1634.63
[3663]	valid_0's rmse: 1634.61
[3664]	valid_0's rmse: 1634.58
[3665]	valid_0's rmse: 1634.57
[3666]	valid_0's rmse: 1634.59
[3667]	valid_0's rmse: 1634.52
[3668]	valid_0's rmse: 1634.51
[3669]	valid_0's rmse: 1634.49
[3670]	valid_0's rmse: 1634.46
[3671]	valid_0's rmse: 1634.43
[3672]	valid_0's rmse: 1634.46
[3673]	valid_0's rmse: 1634.4
[3674]	valid_0's rmse: 1634.41
[3675]	valid_0's rmse: 1634.42
[3676]	valid_0's rmse: 1634.39
[3677]	valid_0's rmse: 1634.4
[3678]	valid_0's rmse: 1634.42
[3679]	valid_0's rmse: 1634.4
[3680]	valid_0's rmse: 1634.38
[3681]	valid_0's rmse: 1634.4
[3682]	valid_0's rmse: 1634.39
[3683]	valid_0's rmse: 1634.38
[3684]	valid_0's rmse: 1634.38
[3685]	valid_0's rmse: 1634.34
[3686]	valid_0's rmse: 1634.35
[3687]	valid_0's rmse: 1634.36
[3688]	valid_0's rmse: 1634.4
[3689]	valid_0's rmse: 1634.41
[3690]	valid_0's rmse: 1634.41
[3691]	valid_0's rmse: 1634.4
[3692]	valid_0's rmse: 1634.37
[3693]	valid_0's rmse: 1634.38
[3694]	valid_0's rmse: 1634.4
[3695]	valid_0's rmse: 1634.34
[3696]	valid_0's rmse: 1634.33
[3697]	valid_0's rmse: 1634.36
[3698]	valid_0's rmse: 1634.35
[3699]	valid_0's rmse: 1634.36
[3700]	valid_0's rmse: 1634.33
[3701]	valid_0's rmse: 1634.34
[3702]	valid_0's rmse: 1634.39
[3703]	valid_0's rmse: 1634.42
[3704]	valid_0's rmse: 1634.44
[3705]	valid_0's rmse: 1634.43
[3706]	valid_0's rmse: 1634.41
[3707]	valid_0's rmse: 1634.42
[3708]	valid_0's rmse: 1634.43
[3709]	valid_0's rmse: 1634.45
[3710]	valid_0's rmse: 1634.45
[3711]	valid_0's rmse: 1634.42
[3712]	valid_0's rmse: 1634.41
[3713]	valid_0's rmse: 1634.4
[3714]	valid_0's rmse: 1634.42
[3715]	valid_0's rmse: 1634.43
[3716]	valid_0's rmse: 1634.42
[3717]	valid_0's rmse: 1634.36
[3718]	valid_0's rmse: 1634.34
[3719]	valid_0's rmse: 1634.32
[3720]	valid_0's rmse: 1634.3
[3721]	valid_0's rmse: 1634.31
[3722]	valid_0's rmse: 1634.25
[3723]	valid_0's rmse: 1634.24
[3724]	valid_0's rmse: 1634.27
[3725]	valid_0's rmse: 1634.26
[3726]	valid_0's rmse: 1634.27
[3727]	valid_0's rmse: 1634.27
[3728]	valid_0's rmse: 1634.28
[3729]	valid_0's rmse: 1634.3
[3730]	valid_0's rmse: 1634.31
[3731]	valid_0's rmse: 1634.33
[3732]	valid_0's rmse: 1634.31
[3733]	valid_0's rmse: 1634.3
[3734]	valid_0's rmse: 1634.31
[3735]	valid_0's rmse: 1634.3
[3736]	valid_0's rmse: 1634.31
[3737]	valid_0's rmse: 1634.29
[3738]	valid_0's rmse: 1634.29
[3739]	valid_0's rmse: 1634.31
[3740]	valid_0's rmse: 1634.31
[3741]	valid_0's rmse: 1634.33
[3742]	valid_0's rmse: 1634.32
[3743]	valid_0's rmse: 1634.32
[3744]	valid_0's rmse: 1634.28
[3745]	valid_0's rmse: 1634.28
[3746]	valid_0's rmse: 1634.27
[3747]	valid_0's rmse: 1634.25
[3748]	valid_0's rmse: 1634.22
[3749]	valid_0's rmse: 1634.18
[3750]	valid_0's rmse: 1634.15
[3751]	valid_0's rmse: 1634.19
[3752]	valid_0's rmse: 1634.18
[3753]	valid_0's rmse: 1634.17
[3754]	valid_0's rmse: 1634.18
[3755]	valid_0's rmse: 1634.2
[3756]	valid_0's rmse: 1634.18
[3757]	valid_0's rmse: 1634.14
[3758]	valid_0's rmse: 1634.15
[3759]	valid_0's rmse: 1634.16
[3760]	valid_0's rmse: 1634.11
[3761]	valid_0's rmse: 1634.11
[3762]	valid_0's rmse: 1634.1
[3763]	valid_0's rmse: 1634.11
[3764]	valid_0's rmse: 1634.1
[3765]	valid_0's rmse: 1634.08
[3766]	valid_0's rmse: 1634.08
[3767]	valid_0's rmse: 1634.08
[3768]	valid_0's rmse: 1634.08
[3769]	valid_0's rmse: 1634.1
[3770]	valid_0's rmse: 1634.08
[3771]	valid_0's rmse: 1634.08
[3772]	valid_0's rmse: 1634.06
[3773]	valid_0's rmse: 1634.02
[3774]	valid_0's rmse: 1633.97
[3775]	valid_0's rmse: 1633.96
[3776]	valid_0's rmse: 1633.96
[3777]	valid_0's rmse: 1633.94
[3778]	valid_0's rmse: 1633.94
[3779]	valid_0's rmse: 1633.92
[3780]	valid_0's rmse: 1633.93
[3781]	valid_0's rmse: 1633.96
[3782]	valid_0's rmse: 1633.94
[3783]	valid_0's rmse: 1633.93
[3784]	valid_0's rmse: 1633.95
[3785]	valid_0's rmse: 1633.91
[3786]	valid_0's rmse: 1633.91
[3787]	valid_0's rmse: 1633.91
[3788]	valid_0's rmse: 1633.9
[3789]	valid_0's rmse: 1633.95
[3790]	valid_0's rmse: 1633.94
[3791]	valid_0's rmse: 1633.93
[3792]	valid_0's rmse: 1633.88
[3793]	valid_0's rmse: 1633.87
[3794]	valid_0's rmse: 1633.9
[3795]	valid_0's rmse: 1633.9
[3796]	valid_0's rmse: 1633.92
[3797]	valid_0's rmse: 1633.92
[3798]	valid_0's rmse: 1633.89
[3799]	valid_0's rmse: 1633.88
[3800]	valid_0's rmse: 1633.87
[3801]	valid_0's rmse: 1633.86
[3802]	valid_0's rmse: 1633.87
[3803]	valid_0's rmse: 1633.89
[3804]	valid_0's rmse: 1633.9
[3805]	valid_0's rmse: 1633.89
[3806]	valid_0's rmse: 1633.88
[3807]	valid_0's rmse: 1633.9
[3808]	valid_0's rmse: 1633.88
[3809]	valid_0's rmse: 1633.89
[3810]	valid_0's rmse: 1633.88
[3811]	valid_0's rmse: 1633.9
[3812]	valid_0's rmse: 1633.9
[3813]	valid_0's rmse: 1633.85
[3814]	valid_0's rmse: 1633.88
[3815]	valid_0's rmse: 1633.89
[3816]	valid_0's rmse: 1633.86
[3817]	valid_0's rmse: 1633.86
[3818]	valid_0's rmse: 1633.86
[3819]	valid_0's rmse: 1633.86
[3820]	valid_0's rmse: 1633.86
[3821]	valid_0's rmse: 1633.92
[3822]	valid_0's rmse: 1633.92
[3823]	valid_0's rmse: 1633.9
[3824]	valid_0's rmse: 1633.9
[3825]	valid_0's rmse: 1633.9
[3826]	valid_0's rmse: 1633.9
[3827]	valid_0's rmse: 1633.91
[3828]	valid_0's rmse: 1633.88
[3829]	valid_0's rmse: 1633.86
[3830]	valid_0's rmse: 1633.87
[3831]	valid_0's rmse: 1633.88
[3832]	valid_0's rmse: 1633.87
[3833]	valid_0's rmse: 1633.86
[3834]	valid_0's rmse: 1633.88
[3835]	valid_0's rmse: 1633.87
[3836]	valid_0's rmse: 1633.87
[3837]	valid_0's rmse: 1633.86
[3838]	valid_0's rmse: 1633.84
[3839]	valid_0's rmse: 1633.84
[3840]	valid_0's rmse: 1633.86
[3841]	valid_0's rmse: 1633.82
[3842]	valid_0's rmse: 1633.83
[3843]	valid_0's rmse: 1633.78
[3844]	valid_0's rmse: 1633.77
[3845]	valid_0's rmse: 1633.72
[3846]	valid_0's rmse: 1633.72
[3847]	valid_0's rmse: 1633.71
[3848]	valid_0's rmse: 1633.69
[3849]	valid_0's rmse: 1633.7
[3850]	valid_0's rmse: 1633.67
[3851]	valid_0's rmse: 1633.64
[3852]	valid_0's rmse: 1633.64
[3853]	valid_0's rmse: 1633.63
[3854]	valid_0's rmse: 1633.63
[3855]	valid_0's rmse: 1633.64
[3856]	valid_0's rmse: 1633.66
[3857]	valid_0's rmse: 1633.7
[3858]	valid_0's rmse: 1633.68
[3859]	valid_0's rmse: 1633.68
[3860]	valid_0's rmse: 1633.68
[3861]	valid_0's rmse: 1633.72
[3862]	valid_0's rmse: 1633.69
[3863]	valid_0's rmse: 1633.69
[3864]	valid_0's rmse: 1633.69
[3865]	valid_0's rmse: 1633.68
[3866]	valid_0's rmse: 1633.65
[3867]	valid_0's rmse: 1633.61
[3868]	valid_0's rmse: 1633.57
[3869]	valid_0's rmse: 1633.54
[3870]	valid_0's rmse: 1633.54
[3871]	valid_0's rmse: 1633.52
[3872]	valid_0's rmse: 1633.54
[3873]	valid_0's rmse: 1633.52
[3874]	valid_0's rmse: 1633.47
[3875]	valid_0's rmse: 1633.42
[3876]	valid_0's rmse: 1633.43
[3877]	valid_0's rmse: 1633.44
[3878]	valid_0's rmse: 1633.42
[3879]	valid_0's rmse: 1633.43
[3880]	valid_0's rmse: 1633.43
[3881]	valid_0's rmse: 1633.37
[3882]	valid_0's rmse: 1633.34
[3883]	valid_0's rmse: 1633.28
[3884]	valid_0's rmse: 1633.3
[3885]	valid_0's rmse: 1633.28
[3886]	valid_0's rmse: 1633.29
[3887]	valid_0's rmse: 1633.3
[3888]	valid_0's rmse: 1633.26
[3889]	valid_0's rmse: 1633.26
[3890]	valid_0's rmse: 1633.26
[3891]	valid_0's rmse: 1633.26
[3892]	valid_0's rmse: 1633.21
[3893]	valid_0's rmse: 1633.19
[3894]	valid_0's rmse: 1633.16
[3895]	valid_0's rmse: 1633.17
[3896]	valid_0's rmse: 1633.18
[3897]	valid_0's rmse: 1633.23
[3898]	valid_0's rmse: 1633.29
[3899]	valid_0's rmse: 1633.31
[3900]	valid_0's rmse: 1633.3
[3901]	valid_0's rmse: 1633.32
[3902]	valid_0's rmse: 1633.32
[3903]	valid_0's rmse: 1633.32
[3904]	valid_0's rmse: 1633.35
[3905]	valid_0's rmse: 1633.36
[3906]	valid_0's rmse: 1633.36
[3907]	valid_0's rmse: 1633.36
[3908]	valid_0's rmse: 1633.37
[3909]	valid_0's rmse: 1633.39
[3910]	valid_0's rmse: 1633.39
[3911]	valid_0's rmse: 1633.36
[3912]	valid_0's rmse: 1633.39
[3913]	valid_0's rmse: 1633.42
[3914]	valid_0's rmse: 1633.41
[3915]	valid_0's rmse: 1633.41
[3916]	valid_0's rmse: 1633.42
[3917]	valid_0's rmse: 1633.43
[3918]	valid_0's rmse: 1633.44
[3919]	valid_0's rmse: 1633.45
[3920]	valid_0's rmse: 1633.41
[3921]	valid_0's rmse: 1633.4
[3922]	valid_0's rmse: 1633.39
[3923]	valid_0's rmse: 1633.4
[3924]	valid_0's rmse: 1633.39
[3925]	valid_0's rmse: 1633.4
[3926]	valid_0's rmse: 1633.39
[3927]	valid_0's rmse: 1633.38
[3928]	valid_0's rmse: 1633.39
[3929]	valid_0's rmse: 1633.41
[3930]	valid_0's rmse: 1633.4
[3931]	valid_0's rmse: 1633.42
[3932]	valid_0's rmse: 1633.43
[3933]	valid_0's rmse: 1633.44
[3934]	valid_0's rmse: 1633.42
[3935]	valid_0's rmse: 1633.41
[3936]	valid_0's rmse: 1633.42
[3937]	valid_0's rmse: 1633.43
[3938]	valid_0's rmse: 1633.42
[3939]	valid_0's rmse: 1633.4
[3940]	valid_0's rmse: 1633.4
[3941]	valid_0's rmse: 1633.41
[3942]	valid_0's rmse: 1633.41
[3943]	valid_0's rmse: 1633.38
[3944]	valid_0's rmse: 1633.36
[3945]	valid_0's rmse: 1633.38
[3946]	valid_0's rmse: 1633.38
[3947]	valid_0's rmse: 1633.38
[3948]	valid_0's rmse: 1633.4
[3949]	valid_0's rmse: 1633.37
[3950]	valid_0's rmse: 1633.32
[3951]	valid_0's rmse: 1633.37
[3952]	valid_0's rmse: 1633.32
[3953]	valid_0's rmse: 1633.28
[3954]	valid_0's rmse: 1633.27
[3955]	valid_0's rmse: 1633.33
[3956]	valid_0's rmse: 1633.3
[3957]	valid_0's rmse: 1633.31
[3958]	valid_0's rmse: 1633.29
[3959]	valid_0's rmse: 1633.32
[3960]	valid_0's rmse: 1633.38
[3961]	valid_0's rmse: 1633.35
[3962]	valid_0's rmse: 1633.35
[3963]	valid_0's rmse: 1633.35
[3964]	valid_0's rmse: 1633.43
[3965]	valid_0's rmse: 1633.42
[3966]	valid_0's rmse: 1633.43
[3967]	valid_0's rmse: 1633.45
[3968]	valid_0's rmse: 1633.47
[3969]	valid_0's rmse: 1633.47
[3970]	valid_0's rmse: 1633.51
[3971]	valid_0's rmse: 1633.54
[3972]	valid_0's rmse: 1633.53
[3973]	valid_0's rmse: 1633.54
[3974]	valid_0's rmse: 1633.52
[3975]	valid_0's rmse: 1633.53
[3976]	valid_0's rmse: 1633.53
[3977]	valid_0's rmse: 1633.49
[3978]	valid_0's rmse: 1633.48
[3979]	valid_0's rmse: 1633.48
[3980]	valid_0's rmse: 1633.49
[3981]	valid_0's rmse: 1633.49
[3982]	valid_0's rmse: 1633.5
[3983]	valid_0's rmse: 1633.47
[3984]	valid_0's rmse: 1633.49
[3985]	valid_0's rmse: 1633.5
[3986]	valid_0's rmse: 1633.49
[3987]	valid_0's rmse: 1633.46
[3988]	valid_0's rmse: 1633.49
[3989]	valid_0's rmse: 1633.51
[3990]	valid_0's rmse: 1633.5
[3991]	valid_0's rmse: 1633.52
[3992]	valid_0's rmse: 1633.58
[3993]	valid_0's rmse: 1633.58
[3994]	valid_0's rmse: 1633.58
[3995]	valid_0's rmse: 1633.59
[3996]	valid_0's rmse: 1633.54
[3997]	valid_0's rmse: 1633.56
[3998]	valid_0's rmse: 1633.54
[3999]	valid_0's rmse: 1633.53
[4000]	valid_0's rmse: 1633.53
[4001]	valid_0's rmse: 1633.58
[4002]	valid_0's rmse: 1633.59
[4003]	valid_0's rmse: 1633.63
[4004]	valid_0's rmse: 1633.62
[4005]	valid_0's rmse: 1633.62
[4006]	valid_0's rmse: 1633.63
[4007]	valid_0's rmse: 1633.62
[4008]	valid_0's rmse: 1633.6
[4009]	valid_0's rmse: 1633.59
[4010]	valid_0's rmse: 1633.6
[4011]	valid_0's rmse: 1633.61
[4012]	valid_0's rmse: 1633.62
[4013]	valid_0's rmse: 1633.59
[4014]	valid_0's rmse: 1633.6
[4015]	valid_0's rmse: 1633.57
[4016]	valid_0's rmse: 1633.57
[4017]	valid_0's rmse: 1633.57
[4018]	valid_0's rmse: 1633.59
[4019]	valid_0's rmse: 1633.57
[4020]	valid_0's rmse: 1633.58
[4021]	valid_0's rmse: 1633.56
[4022]	valid_0's rmse: 1633.57
[4023]	valid_0's rmse: 1633.57
[4024]	valid_0's rmse: 1633.58
[4025]	valid_0's rmse: 1633.58
[4026]	valid_0's rmse: 1633.55
[4027]	valid_0's rmse: 1633.55
[4028]	valid_0's rmse: 1633.55
[4029]	valid_0's rmse: 1633.53
[4030]	valid_0's rmse: 1633.56
[4031]	valid_0's rmse: 1633.55
[4032]	valid_0's rmse: 1633.61
[4033]	valid_0's rmse: 1633.59
[4034]	valid_0's rmse: 1633.59
[4035]	valid_0's rmse: 1633.62
[4036]	valid_0's rmse: 1633.66
[4037]	valid_0's rmse: 1633.65
[4038]	valid_0's rmse: 1633.7
[4039]	valid_0's rmse: 1633.71
[4040]	valid_0's rmse: 1633.73
[4041]	valid_0's rmse: 1633.71
[4042]	valid_0's rmse: 1633.75
[4043]	valid_0's rmse: 1633.75
[4044]	valid_0's rmse: 1633.74
[4045]	valid_0's rmse: 1633.75
[4046]	valid_0's rmse: 1633.75
[4047]	valid_0's rmse: 1633.75
[4048]	valid_0's rmse: 1633.79
[4049]	valid_0's rmse: 1633.78
[4050]	valid_0's rmse: 1633.76
[4051]	valid_0's rmse: 1633.79
[4052]	valid_0's rmse: 1633.81
[4053]	valid_0's rmse: 1633.81
[4054]	valid_0's rmse: 1633.81
[4055]	valid_0's rmse: 1633.86
[4056]	valid_0's rmse: 1633.83
[4057]	valid_0's rmse: 1633.83
[4058]	valid_0's rmse: 1633.83
[4059]	valid_0's rmse: 1633.87
[4060]	valid_0's rmse: 1633.88
[4061]	valid_0's rmse: 1633.83
[4062]	valid_0's rmse: 1633.82
[4063]	valid_0's rmse: 1633.85
[4064]	valid_0's rmse: 1633.84
[4065]	valid_0's rmse: 1633.84
[4066]	valid_0's rmse: 1633.85
[4067]	valid_0's rmse: 1633.88
[4068]	valid_0's rmse: 1633.94
[4069]	valid_0's rmse: 1633.92
[4070]	valid_0's rmse: 1633.9
[4071]	valid_0's rmse: 1633.91
[4072]	valid_0's rmse: 1633.9
[4073]	valid_0's rmse: 1633.92
[4074]	valid_0's rmse: 1633.9
[4075]	valid_0's rmse: 1633.88
[4076]	valid_0's rmse: 1633.86
[4077]	valid_0's rmse: 1633.9
[4078]	valid_0's rmse: 1633.88
[4079]	valid_0's rmse: 1633.85
[4080]	valid_0's rmse: 1633.88
[4081]	valid_0's rmse: 1633.86
[4082]	valid_0's rmse: 1633.85
[4083]	valid_0's rmse: 1633.84
[4084]	valid_0's rmse: 1633.85
[4085]	valid_0's rmse: 1633.83
[4086]	valid_0's rmse: 1633.82
[4087]	valid_0's rmse: 1633.81
[4088]	valid_0's rmse: 1633.81
[4089]	valid_0's rmse: 1633.83
[4090]	valid_0's rmse: 1633.84
[4091]	valid_0's rmse: 1633.85
[4092]	valid_0's rmse: 1633.84
[4093]	valid_0's rmse: 1633.83
[4094]	valid_0's rmse: 1633.81
[4095]	valid_0's rmse: 1633.82
[4096]	valid_0's rmse: 1633.82
[4097]	valid_0's rmse: 1633.86
[4098]	valid_0's rmse: 1633.84
[4099]	valid_0's rmse: 1633.83
[4100]	valid_0's rmse: 1633.83
[4101]	valid_0's rmse: 1633.85
[4102]	valid_0's rmse: 1633.81
[4103]	valid_0's rmse: 1633.8
[4104]	valid_0's rmse: 1633.79
[4105]	valid_0's rmse: 1633.79
[4106]	valid_0's rmse: 1633.77
[4107]	valid_0's rmse: 1633.78
[4108]	valid_0's rmse: 1633.77
[4109]	valid_0's rmse: 1633.77
[4110]	valid_0's rmse: 1633.78
[4111]	valid_0's rmse: 1633.75
[4112]	valid_0's rmse: 1633.76
[4113]	valid_0's rmse: 1633.76
[4114]	valid_0's rmse: 1633.77
[4115]	valid_0's rmse: 1633.76
[4116]	valid_0's rmse: 1633.74
[4117]	valid_0's rmse: 1633.75
[4118]	valid_0's rmse: 1633.77
[4119]	valid_0's rmse: 1633.77
[4120]	valid_0's rmse: 1633.76
[4121]	valid_0's rmse: 1633.75
[4122]	valid_0's rmse: 1633.77
[4123]	valid_0's rmse: 1633.75
[4124]	valid_0's rmse: 1633.74
[4125]	valid_0's rmse: 1633.73
[4126]	valid_0's rmse: 1633.79
[4127]	valid_0's rmse: 1633.78
[4128]	valid_0's rmse: 1633.75
[4129]	valid_0's rmse: 1633.75
[4130]	valid_0's rmse: 1633.78
[4131]	valid_0's rmse: 1633.79
[4132]	valid_0's rmse: 1633.78
[4133]	valid_0's rmse: 1633.79
[4134]	valid_0's rmse: 1633.77
[4135]	valid_0's rmse: 1633.74
[4136]	valid_0's rmse: 1633.72
[4137]	valid_0's rmse: 1633.72
[4138]	valid_0's rmse: 1633.75
[4139]	valid_0's rmse: 1633.74
[4140]	valid_0's rmse: 1633.73
[4141]	valid_0's rmse: 1633.73
[4142]	valid_0's rmse: 1633.71
[4143]	valid_0's rmse: 1633.71
[4144]	valid_0's rmse: 1633.67
[4145]	valid_0's rmse: 1633.68
[4146]	valid_0's rmse: 1633.69
[4147]	valid_0's rmse: 1633.68
[4148]	valid_0's rmse: 1633.68
[4149]	valid_0's rmse: 1633.71
[4150]	valid_0's rmse: 1633.7
[4151]	valid_0's rmse: 1633.72
[4152]	valid_0's rmse: 1633.74
[4153]	valid_0's rmse: 1633.75
[4154]	valid_0's rmse: 1633.77
[4155]	valid_0's rmse: 1633.78
[4156]	valid_0's rmse: 1633.78
[4157]	valid_0's rmse: 1633.81
[4158]	valid_0's rmse: 1633.83
[4159]	valid_0's rmse: 1633.83
[4160]	valid_0's rmse: 1633.82
[4161]	valid_0's rmse: 1633.78
[4162]	valid_0's rmse: 1633.76
[4163]	valid_0's rmse: 1633.77
[4164]	valid_0's rmse: 1633.76
[4165]	valid_0's rmse: 1633.77
[4166]	valid_0's rmse: 1633.77
[4167]	valid_0's rmse: 1633.77
[4168]	valid_0's rmse: 1633.76
[4169]	valid_0's rmse: 1633.73
[4170]	valid_0's rmse: 1633.75
[4171]	valid_0's rmse: 1633.76
[4172]	valid_0's rmse: 1633.76
[4173]	valid_0's rmse: 1633.75
[4174]	valid_0's rmse: 1633.77
[4175]	valid_0's rmse: 1633.79
[4176]	valid_0's rmse: 1633.79
[4177]	valid_0's rmse: 1633.82
[4178]	valid_0's rmse: 1633.82
[4179]	valid_0's rmse: 1633.81
[4180]	valid_0's rmse: 1633.81
[4181]	valid_0's rmse: 1633.81
[4182]	valid_0's rmse: 1633.8
[4183]	valid_0's rmse: 1633.79
[4184]	valid_0's rmse: 1633.81
[4185]	valid_0's rmse: 1633.83
[4186]	valid_0's rmse: 1633.84
[4187]	valid_0's rmse: 1633.88
[4188]	valid_0's rmse: 1633.87
[4189]	valid_0's rmse: 1633.9
[4190]	valid_0's rmse: 1633.93
[4191]	valid_0's rmse: 1633.93
[4192]	valid_0's rmse: 1633.91
[4193]	valid_0's rmse: 1633.94
[4194]	valid_0's rmse: 1633.98
[4195]	valid_0's rmse: 1633.97
[4196]	valid_0's rmse: 1633.94
[4197]	valid_0's rmse: 1633.93
[4198]	valid_0's rmse: 1633.93
[4199]	valid_0's rmse: 1633.88
[4200]	valid_0's rmse: 1633.88
[4201]	valid_0's rmse: 1633.87
[4202]	valid_0's rmse: 1633.89
[4203]	valid_0's rmse: 1633.87
[4204]	valid_0's rmse: 1633.87
[4205]	valid_0's rmse: 1633.85
[4206]	valid_0's rmse: 1633.85
[4207]	valid_0's rmse: 1633.85
[4208]	valid_0's rmse: 1633.86
[4209]	valid_0's rmse: 1633.88
[4210]	valid_0's rmse: 1633.89
[4211]	valid_0's rmse: 1633.88
[4212]	valid_0's rmse: 1633.91
[4213]	valid_0's rmse: 1633.87
[4214]	valid_0's rmse: 1633.88
[4215]	valid_0's rmse: 1633.87
[4216]	valid_0's rmse: 1633.84
[4217]	valid_0's rmse: 1633.85
[4218]	valid_0's rmse: 1633.85
[4219]	valid_0's rmse: 1633.84
[4220]	valid_0's rmse: 1633.84
[4221]	valid_0's rmse: 1633.79
[4222]	valid_0's rmse: 1633.77
[4223]	valid_0's rmse: 1633.76
[4224]	valid_0's rmse: 1633.76
[4225]	valid_0's rmse: 1633.74
[4226]	valid_0's rmse: 1633.72
[4227]	valid_0's rmse: 1633.73
[4228]	valid_0's rmse: 1633.74
[4229]	valid_0's rmse: 1633.72
[4230]	valid_0's rmse: 1633.72
[4231]	valid_0's rmse: 1633.72
[4232]	valid_0's rmse: 1633.73
[4233]	valid_0's rmse: 1633.7
[4234]	valid_0's rmse: 1633.72
[4235]	valid_0's rmse: 1633.7
[4236]	valid_0's rmse: 1633.71
[4237]	valid_0's rmse: 1633.69
[4238]	valid_0's rmse: 1633.7
[4239]	valid_0's rmse: 1633.68
[4240]	valid_0's rmse: 1633.67
[4241]	valid_0's rmse: 1633.67
[4242]	valid_0's rmse: 1633.68
[4243]	valid_0's rmse: 1633.68
[4244]	valid_0's rmse: 1633.67
[4245]	valid_0's rmse: 1633.64
[4246]	valid_0's rmse: 1633.64
[4247]	valid_0's rmse: 1633.67
[4248]	valid_0's rmse: 1633.68
[4249]	valid_0's rmse: 1633.67
[4250]	valid_0's rmse: 1633.63
[4251]	valid_0's rmse: 1633.61
[4252]	valid_0's rmse: 1633.61
[4253]	valid_0's rmse: 1633.61
[4254]	valid_0's rmse: 1633.64
[4255]	valid_0's rmse: 1633.64
[4256]	valid_0's rmse: 1633.63
[4257]	valid_0's rmse: 1633.66
[4258]	valid_0's rmse: 1633.7
[4259]	valid_0's rmse: 1633.72
[4260]	valid_0's rmse: 1633.71
[4261]	valid_0's rmse: 1633.74
[4262]	valid_0's rmse: 1633.76
[4263]	valid_0's rmse: 1633.76
[4264]	valid_0's rmse: 1633.82
[4265]	valid_0's rmse: 1633.82
[4266]	valid_0's rmse: 1633.83
[4267]	valid_0's rmse: 1633.81
[4268]	valid_0's rmse: 1633.82
[4269]	valid_0's rmse: 1633.84
[4270]	valid_0's rmse: 1633.84
[4271]	valid_0's rmse: 1633.83
[4272]	valid_0's rmse: 1633.85
[4273]	valid_0's rmse: 1633.84
[4274]	valid_0's rmse: 1633.84
[4275]	valid_0's rmse: 1633.83
[4276]	valid_0's rmse: 1633.83
[4277]	valid_0's rmse: 1633.83
[4278]	valid_0's rmse: 1633.85
[4279]	valid_0's rmse: 1633.87
[4280]	valid_0's rmse: 1633.87
[4281]	valid_0's rmse: 1633.82
[4282]	valid_0's rmse: 1633.84
[4283]	valid_0's rmse: 1633.84
[4284]	valid_0's rmse: 1633.86
[4285]	valid_0's rmse: 1633.85
[4286]	valid_0's rmse: 1633.86
[4287]	valid_0's rmse: 1633.85
[4288]	valid_0's rmse: 1633.83
[4289]	valid_0's rmse: 1633.82
[4290]	valid_0's rmse: 1633.81
[4291]	valid_0's rmse: 1633.82
[4292]	valid_0's rmse: 1633.83
[4293]	valid_0's rmse: 1633.81
[4294]	valid_0's rmse: 1633.84
[4295]	valid_0's rmse: 1633.83
[4296]	valid_0's rmse: 1633.86
[4297]	valid_0's rmse: 1633.86
[4298]	valid_0's rmse: 1633.85
[4299]	valid_0's rmse: 1633.87
[4300]	valid_0's rmse: 1633.89
[4301]	valid_0's rmse: 1633.9
[4302]	valid_0's rmse: 1633.91
[4303]	valid_0's rmse: 1633.89
[4304]	valid_0's rmse: 1633.85
[4305]	valid_0's rmse: 1633.85
[4306]	valid_0's rmse: 1633.86
[4307]	valid_0's rmse: 1633.88
[4308]	valid_0's rmse: 1633.92
[4309]	valid_0's rmse: 1633.93
[4310]	valid_0's rmse: 1633.95
[4311]	valid_0's rmse: 1633.91
[4312]	valid_0's rmse: 1633.89
[4313]	valid_0's rmse: 1633.83
[4314]	valid_0's rmse: 1633.82
[4315]	valid_0's rmse: 1633.79
[4316]	valid_0's rmse: 1633.75
[4317]	valid_0's rmse: 1633.73
[4318]	valid_0's rmse: 1633.77
[4319]	valid_0's rmse: 1633.77
[4320]	valid_0's rmse: 1633.74
[4321]	valid_0's rmse: 1633.73
[4322]	valid_0's rmse: 1633.72
[4323]	valid_0's rmse: 1633.71
[4324]	valid_0's rmse: 1633.71
[4325]	valid_0's rmse: 1633.71
[4326]	valid_0's rmse: 1633.7
[4327]	valid_0's rmse: 1633.69
[4328]	valid_0's rmse: 1633.68
[4329]	valid_0's rmse: 1633.7
[4330]	valid_0's rmse: 1633.71
[4331]	valid_0's rmse: 1633.7
[4332]	valid_0's rmse: 1633.72
[4333]	valid_0's rmse: 1633.72
[4334]	valid_0's rmse: 1633.7
[4335]	valid_0's rmse: 1633.72
[4336]	valid_0's rmse: 1633.72
[4337]	valid_0's rmse: 1633.71
[4338]	valid_0's rmse: 1633.69
[4339]	valid_0's rmse: 1633.7
[4340]	valid_0's rmse: 1633.76
[4341]	valid_0's rmse: 1633.77
[4342]	valid_0's rmse: 1633.77
[4343]	valid_0's rmse: 1633.77
[4344]	valid_0's rmse: 1633.78
[4345]	valid_0's rmse: 1633.76
[4346]	valid_0's rmse: 1633.75
[4347]	valid_0's rmse: 1633.75
[4348]	valid_0's rmse: 1633.71
[4349]	valid_0's rmse: 1633.72
[4350]	valid_0's rmse: 1633.71
[4351]	valid_0's rmse: 1633.71
[4352]	valid_0's rmse: 1633.69
[4353]	valid_0's rmse: 1633.69
[4354]	valid_0's rmse: 1633.69
[4355]	valid_0's rmse: 1633.69
[4356]	valid_0's rmse: 1633.69
[4357]	valid_0's rmse: 1633.67
[4358]	valid_0's rmse: 1633.65
[4359]	valid_0's rmse: 1633.65
[4360]	valid_0's rmse: 1633.7
[4361]	valid_0's rmse: 1633.69
[4362]	valid_0's rmse: 1633.7
[4363]	valid_0's rmse: 1633.69
[4364]	valid_0's rmse: 1633.69
[4365]	valid_0's rmse: 1633.69
[4366]	valid_0's rmse: 1633.72
[4367]	valid_0's rmse: 1633.72
[4368]	valid_0's rmse: 1633.66
[4369]	valid_0's rmse: 1633.7
[4370]	valid_0's rmse: 1633.67
[4371]	valid_0's rmse: 1633.66
[4372]	valid_0's rmse: 1633.7
[4373]	valid_0's rmse: 1633.69
[4374]	valid_0's rmse: 1633.68
[4375]	valid_0's rmse: 1633.7
[4376]	valid_0's rmse: 1633.7
[4377]	valid_0's rmse: 1633.69
[4378]	valid_0's rmse: 1633.69
[4379]	valid_0's rmse: 1633.7
[4380]	valid_0's rmse: 1633.67
[4381]	valid_0's rmse: 1633.66
[4382]	valid_0's rmse: 1633.66
[4383]	valid_0's rmse: 1633.66
[4384]	valid_0's rmse: 1633.65
[4385]	valid_0's rmse: 1633.66
[4386]	valid_0's rmse: 1633.61
[4387]	valid_0's rmse: 1633.63
[4388]	valid_0's rmse: 1633.6
[4389]	valid_0's rmse: 1633.61
[4390]	valid_0's rmse: 1633.63
[4391]	valid_0's rmse: 1633.65
[4392]	valid_0's rmse: 1633.68
[4393]	valid_0's rmse: 1633.67
[4394]	valid_0's rmse: 1633.62
Early stopping, best iteration is:
[3894]	valid_0's rmse: 1633.16

RMSE of the lgbm model on the validation set: 1633.1585371342076
In [86]:
# %%timeit
# light gbm prediction
lgbm.predict(X_valid[:2])
Out[86]:
array([2824.93089504, 1220.24599181])
In [87]:
# table of feature importance
feat_imp = [t for t in zip(features, lgbm.feature_importance())]
feat_imp_df = pd.DataFrame(feat_imp, columns=['feature', 'varimp'])
feat_imp_df = feat_imp_df.sort_values('varimp', ascending=False)
feat_imp_df
Out[87]:
feature varimp
10 postal_code 30146
3 power 23438
1 registration_year 20188
6 registration_month 13430
4 model 10209
5 mileage 9159
8 brand 3402
0 vehicle_type 3188
2 gearbox 2582
9 not_repaired 739
7 fuel_type 339

Catboost¶

In [88]:
# fill missing values as string NaN
df.fillna('NaN', inplace=True)
In [89]:
# %%timeit
# catboost
X1 = df.drop(['price','year_crawled', 'month_crawled',
       'day_crawled', 'time_crawled'], axis=1)
y1 = df['price']


X1_train, X1_test, y1_train, y1_test = train_test_split(
    X1, y1, test_size=0.25, random_state=19)

X1_train, X1_valid, y1_train, y1_valid = train_test_split(
    X1_train, y1_train, test_size=0.20, random_state=19)

# fill missing values as string NaN

cat_features = [
    'vehicle_type',
    'gearbox',
    'model',
    'fuel_type', 
    'brand', 
    'not_repaired'
]

catb = CatBoostRegressor(task_type='GPU', loss_function='RMSE', eval_metric='RMSE', iterations=1500, random_seed=19)

catb.fit(X1_train, y1_train, eval_set=(X1_valid, y1_valid), cat_features=cat_features, verbose=100, use_best_model=True, plot=True)

# Make predictions on the test set
predictions_valid_catb = catb.predict(X1_valid)

result = mse(y1_valid, predictions_valid_catb) ** 0.5 # calculate RMSE on validation set
print()
print("Catboost model on the validation set: ")
catb.best_score_

MetricVisualizer(layout=Layout(align_self='stretch', height='500px'))
Learning rate set to 0.097278
0:	learn: 4294.0143420	test: 4312.0688882	best: 4312.0688882 (0)	total: 125ms	remaining: 3m 7s
100:	learn: 1850.7068059	test: 1865.4927636	best: 1865.4927636 (100)	total: 11.1s	remaining: 2m 33s
200:	learn: 1772.7411209	test: 1799.9096708	best: 1799.9096708 (200)	total: 21.8s	remaining: 2m 20s
300:	learn: 1736.0036154	test: 1775.2809012	best: 1775.2809012 (300)	total: 32.2s	remaining: 2m 8s
400:	learn: 1711.0098842	test: 1759.1834077	best: 1759.1834077 (400)	total: 43.4s	remaining: 1m 58s
500:	learn: 1691.8484583	test: 1749.0225927	best: 1749.0161440 (499)	total: 56.3s	remaining: 1m 52s
600:	learn: 1674.9005953	test: 1740.0733097	best: 1740.0733097 (600)	total: 1m 8s	remaining: 1m 43s
700:	learn: 1661.0412251	test: 1734.2242451	best: 1734.2242451 (700)	total: 1m 21s	remaining: 1m 32s
800:	learn: 1646.8603198	test: 1727.0059576	best: 1726.9695270 (798)	total: 1m 34s	remaining: 1m 22s
900:	learn: 1634.0056510	test: 1722.4670428	best: 1722.4670428 (900)	total: 1m 45s	remaining: 1m 10s
1000:	learn: 1623.3198191	test: 1718.4793385	best: 1718.4793385 (1000)	total: 1m 59s	remaining: 59.7s
1100:	learn: 1614.7202606	test: 1715.9938970	best: 1715.9706353 (1099)	total: 2m 14s	remaining: 48.9s
1200:	learn: 1603.9303403	test: 1711.9758283	best: 1711.9758283 (1200)	total: 2m 27s	remaining: 36.8s
1300:	learn: 1594.9271146	test: 1708.6188187	best: 1708.6188187 (1300)	total: 2m 41s	remaining: 24.6s
1400:	learn: 1587.0520286	test: 1705.6079899	best: 1705.6079899 (1400)	total: 2m 54s	remaining: 12.3s
1499:	learn: 1579.7353030	test: 1703.4617586	best: 1703.4617586 (1499)	total: 3m 8s	remaining: 0us
bestTest = 1703.461759
bestIteration = 1499

Catboost model on the validation set:
Out[89]:
{'learn': {'RMSE': 1579.7353029785438},
 'validation': {'RMSE': 1703.4617585957194}}
In [90]:
# %%timeit
# catboost predictions
catb.predict(new_features)
Out[90]:
array([-1357.82073855,  4612.8681004 ])
In [91]:
# table of feature importance
feat_import = [t for t in zip(features, catb.get_feature_importance())]
feat_import_df = pd.DataFrame(feat_import, columns=['feature', 'varimp'])
feat_import_df = feat_import_df.sort_values('varimp', ascending=False)
feat_import_df
Out[91]:
feature varimp
1 registration_year 36.846183
3 power 24.999310
5 mileage 11.474415
0 vehicle_type 9.598783
8 brand 7.520959
4 model 4.367593
9 not_repaired 1.974238
10 postal_code 1.437106
2 gearbox 0.659555
6 registration_month 0.617220
7 fuel_type 0.504637

Comparing Boosting Models¶

Score¶

In [92]:
# boosting model metrics
data2 = {'xgb': (1754.20, 13.9, 1.98, .0046, 2.67 * 10 ** -4), 'lgb': (1629.89, 66 , 3.75, 0.0113, .012), 'catboost': (1703.46, 277, 11.7, 0.00226, 2.26 * 10 ** -4)}
boost_scores = pd.DataFrame(data=data2, columns=data2.keys(), index=['RMSE', 'train_time', 'train_std', 'pred_time', 'pred_std'])
boost_scores = boost_scores.T
boost_scores
Out[92]:
RMSE train_time train_std pred_time pred_std
xgb 1754.20 13.9 1.98 0.00460 0.000267
lgb 1629.89 66.0 3.75 0.01130 0.012000
catboost 1703.46 277.0 11.70 0.00226 0.000226
In [93]:
# boosting RMSE
px.scatter(boost_scores.RMSE, color_discrete_sequence=[['yellow', 'cyan', 'violet']], labels={'index': 'Model', 'value': 'RMSE Score'}, title='Model Score Comparison', size=boost_scores.RMSE, log_y=True)
xgblgbcatboost16201640166016801700172017401760
variableRMSEModel Score ComparisonModelRMSE Score
plotly-logomark
In [94]:
# comparing train and predict times
fig = go.Figure()
fig.add_trace(go.Bar(
    name='Train',
    x=['xg boost', 'light gbm', 'catboost'], y=boost_scores.train_time,
    error_y=dict(type='data', array=[1.98, 3.75, 11.7])
))
fig.add_trace(go.Bar(
    name='Predict',
    x=['xg boost', 'light gbm', 'catboost'], y=boost_scores.pred_time,
    error_y=dict(type='data', array=[.000267, .012, .000226]
)))

fig.update_layout(barmode='group', title='Model Prediction Time Comparison', template='seaborn')
fig.show()
xg boostlight gbmcatboost050100150200250300
TrainPredictModel Prediction Time Comparison
plotly-logomark
Conclusions¶

Here, we see the lgb model has the lowest RMSE score, while the xgb model has the highest score. These scores can be further lowered by tuning hyperparameters. The catboost model took the longest time training, while the xg boost model took the shortest amount of time. With the predictions, the catboost model took the least amount of time, while the light gbm model took the longest time.

Feature Importance¶

In [95]:
# combining first two feature importance tables
feat1 = xgbr_imp_df.merge(feat_imp_df, on='feature')
final_feat = feat1.merge(feat_import_df, on='feature')
In [96]:
# combining final feature table
final_feat.columns = ['features', 'xgb', 'lgbm', 'catboost']
final_feat
Out[96]:
features xgb lgbm catboost
0 registration_year 0.355426 20188 36.846183
1 power 0.200501 23438 24.999310
2 mileage 0.146589 9159 11.474415
3 not_repaired 0.101944 739 1.974238
4 brand 0.052967 3402 7.520959
5 vehicle_type 0.051839 3188 9.598783
6 gearbox 0.038437 2582 0.659555
7 model 0.021239 10209 4.367593
8 fuel_type 0.012837 339 0.504637
9 registration_month 0.009155 13430 0.617220
10 postal_code 0.009067 30146 1.437106
In [97]:
# normalize importance to same scale
final_feat.lgbm = (final_feat.lgbm - final_feat.lgbm.mean())/ final_feat.lgbm.std()
final_feat.catboost = final_feat.catboost/100
In [98]:
# normalized final feature table
final_feat
Out[98]:
features xgb lgbm catboost
0 registration_year 0.355426 0.946173 0.368462
1 power 0.200501 1.267564 0.249993
2 mileage 0.146589 -0.144477 0.114744
3 not_repaired 0.101944 -0.977126 0.019742
4 brand 0.052967 -0.713783 0.075210
5 vehicle_type 0.051839 -0.734946 0.095988
6 gearbox 0.038437 -0.794873 0.006596
7 model 0.021239 -0.040644 0.043676
8 fuel_type 0.012837 -1.016681 0.005046
9 registration_month 0.009155 0.277879 0.006172
10 postal_code 0.009067 1.930913 0.014371
In [99]:
# feature importance of boosting models
px.bar(final_feat, x='features', y=['xgb', 'lgbm', 'catboost'], barmode='group', template='plotly_dark', title='Feature Importance of Boosting Models', labels={'value': 'Importance'})
registration_yearpowermileagenot_repairedbrandvehicle_typegearboxmodelfuel_typeregistration_monthpostal_code−1−0.500.511.52
variablexgblgbmcatboostFeature Importance of Boosting ModelsfeaturesImportance
plotly-logomark
Conclusion¶

This chart shows the importance of each feature, among the three boosting models. Comparing the lgb model with the others, we see this model placed the most importance on the postal code, while the other models placed very little importance in this feature. The lgb model also placed some importance on registration year and power. The xgb model placed most of the importance on registration year, power, and milage. The catboost model placed most importance on registration year, power, and milage as well.

Comparison between Regression and Boosting¶

In [100]:
# regression scores mean train time
regr_scores.train_time.mean()
Out[100]:
29.197599999999998
In [101]:
# regression scores mean RMSE
regr_scores.RMSE.mean()
Out[101]:
2526.7566666666667
In [102]:
# boost scores mean train time
boost_scores.train_time.mean()
Out[102]:
118.96666666666665
In [103]:
# boost scores mean RMSE
boost_scores.RMSE.mean()
Out[103]:
1695.8500000000001
In [104]:
# model training time comparisons
fig = go.Figure()
fig.add_trace(go.Bar(
    name='regression',
    x=['decision tree', 'random forest', 'linear regression'], y=regr_scores.train_time,
    error_y=dict(type='data', array=[1.5, 10.7, .0691])
))
fig.add_trace(go.Bar(
    name='boost',
    x=['xg boost', 'light gbm', 'catboost'], y=boost_scores.train_time,
    error_y=dict(type='data', array=[1.98, 3.75, 11.7])
))
fig.update_layout(barmode='group', title='Model Training Time Comparison', template='seaborn')
fig.show()
decision treerandom forestlinear regressionxg boostlight gbmcatboost050100150200250300
regressionboostModel Training Time Comparison
plotly-logomark
In [105]:
# comparing prediction times
fig = go.Figure()

fig.add_trace(go.Bar(
    name='regression',
    x=['decision tree', 'random forest', 'linear regression'], y=regr_scores.pred_time,
    error_y=dict(type='data', array=[.00518, .00158, .00301]
)))
fig.add_trace(go.Bar(
    name='boost',
    x=['xg boost', 'light gbm', 'catboost'], y=boost_scores.pred_time,
    error_y=dict(type='data', array=[.000267, .012, .000226]
)))

fig.update_layout(barmode='group', title='Model Prediction Time Comparison', template='seaborn')
fig.show()
decision treerandom forestlinear regressionxg boostlight gbmcatboost00.0050.010.0150.02
regressionboostModel Prediction Time Comparison
plotly-logomark
In [106]:
# comparing RMSE Scores
fig = go.Figure()

fig.add_trace(go.Scatter(
    y=regr_scores.RMSE,
    x=['decision tree', 'random forest', 'linear regression'] )
)
fig.add_trace(go.Scatter(
    y=boost_scores.RMSE, 
    x=['xg boost', 'light gbm', 'catboost'] )
)

fig.update_layout(title='Model RMSE Comparison', template='plotly_dark')
fig.show()
decision treerandom forestlinear regressionxg boostlight gbmcatboost160018002000220024002600280030003200
trace 0trace 1Model RMSE Comparison
plotly-logomark

Final Model¶

Light GBM¶

In [107]:
# Make predictions on the validation set
predictions_test_lgbm = lgbm.predict(X_test)

result = mse(y_test, predictions_test_lgbm) ** 0.5 # calculate RMSE on test set
print()
print("RMSE of the lgbm model on the test set:", result)

RMSE of the lgbm model on the test set: 1637.7809549995363

The final model has an RMSE score of 1639.44 with the test set.

Overall Conclusions¶

The boosting models contain a good balance between speed and model prediction quality. The RMSE scores of the boosting models were better than those of the regression models. Prediction times for either set of models is negligible. Overall, we suggest Rusty Bargain to implement a boosting machine learning model that will predict the market value of a car. From our data, we suggest that they start with a light gbm model, and tune the hyperparameters to achieve a low RMSE. The light gbm model is relatively fast at training, and the predictions are fast. Most importantly, the model prediction quality is the best of the options we've displayed. Another benefit of the light gbm model is its ability to natively handle missing values. This an benefit UX on the app, as users would not have to make up data just to run the model.